Configuring, Upgrading A 6 Core for retouching [Archive]

Honumaui

Aug 21, 2010, 08:33 PM

I do PS and design for a living largest files tend to be about 1 gig or so

the one thing is cache/scratch and working files on the same disc is not as efficient as separating them ? something to think about
especially when writing huge files
since it is basically scratching to the same disc as its writing !

I am not a big fan of long hard work on a Raid 0 ? I use it for some things but not long drawn out work I would rather be working on a safer system should something happen

if I have to redo 4 hours of PS work that means the 4 hours I lost the 4 hours to redo it and the extra hours I push off my next client to get caught back up !
at my $ per hour its cheaper to have more HDs compared to loosing that time

just a thought :)

the 40 gig at $99 thinking of trying them for cache/scratch in raid 0 and use my 100 gig RE for other things
I currently have two 100 gig OWC RE SSD drives one for cache one for boot
areca 1222x with 8 750 drives setup raid 6 for working files and two stand along raid 5 boxes one for bu one for time machine and other offline offsite setups

I might say get a good card and get some externals so you can have fast storage for your working files and not have the chance of going down as much

to me I say if you have a raid 0 and a drive dies ? well its no difference than having a single HD and it dies ? either way you are hosed so I am not against raid 0 as a raid setup I am against betting against me loosing time and work !!!!
I prefer raid 1/0 or even raid 1 or Raid 6 or another form of keep working if a drive dies setup along with some speed
some hate raid 6 here ? but for some of us who deal with lots of large files and dont want the loss of raid 1/0 in disc space vs how many I have to run !
I think Raid 6 on the areca gives me the speed I want with the batter bu module its one more small safety for write issues that could happen

one other setup thought I might say get 4 2TB WD black raid 1/0 using disk utility those go in the sleds
take out my dvd and put it in external and then use those two ports for my SSD one for boot one for cache/scratch
use external BU

sounds like a fine setup but my thoughts for you would be to try comparing yourself having your scratch and working files on the same disc vs separate ? you might find you are OK with them on both

if you get the setup you mentioned I would try to split those two 50 and use one for cache/scratch one for main working files and compare the times

the SSD are a time saver ! this way you can do some testing the setup you wrote and the two drives separate

I have found I take advice then try it make notes compare it to my own findings
I would not take my word or any one else's here as the only thing to do but as good advice to start with and test for yourself ! :)
hope this helps you out

Giuly

Aug 21, 2010, 08:34 PM

Honumaui

Aug 21, 2010, 08:52 PM

You know that the Extreme Pro RE is just a real expensive last-gen SandForce-1200 drive and even OWC doesn't state/know why you should prefer it over the normal Extreme Pro with 60GB for RAID, other than a "RAID-READY Enhanced!!1" from the marketing department?
Only thing about it are the 5 instead of 3 years of warranty.

and 28% Enterprise class over-provisioning VS 7% over-provisioning
in real words %28 vs %7 ?

and their is some who knows if this over-provisioning really works out to the drives benefit ?
in theory with a drive as cache/scratch filling up and emptying out a bunch it might help out ?

so I do agree it might depend on use to decide on which edition to get and how its use is ?

and the 40s at $99 two of them in raid 0 would be better for cache/scratch than a single 100 IMHO :) and my plans to do this week ! is add two of these to my system and move it off the single SSD for scratch
using two SSD in scratch/cache I did gain a very small bump in some things and the fact $198 for 80 gigs is not to much

JulianBoolean

Aug 21, 2010, 09:43 PM

the one thing is cache/scratch and working files on the same disc is not as efficient as separating them ? something to think about especially when writing huge files since it is basically scratching to the same disc as its writing ! .....continued

Thanks for the detailed response Honumaui, much appreciated! :) I've always had my scratch disk on its own, but I'm considering putting my scratch and only my currently working files on the same two disk ssd array with Raid 0 because:

A. Could be wrong but, I don't think photoshop is ever accessing (writing) to the scratch drive and saving the current file at the same time, because saving your file, is a one thread deal. Can't do anything else in PS while it's saving. I need really fast photoshop incremental saves (writes) and 2 SSDs seems like a good fit for that concern.

B. Because my currently working image can demand disk space ranging anywhere from 400 megs to 25 gigs, it's a way to have some flexibility as to how much scratch space I have available, vs. having fixed partitions or disk sizes. Photoshop will just use whatever is left from over for scratch. I don't think fragmentation will be an issue will at least 75 gigs of free space at all times.

C. I will get 2x the speed of an allready fast SSD for both tasks

As for the increased danger of data loss with Raid 0, I'm thinking that the external FW drive with time machine will save my bacon with auto saves every hour on the hour.

Thoughts?, Counterpoints? -JB

Honumaui

Aug 22, 2010, 12:36 AM

at least what I have tested when writing MY large files it hits the scratch

I have most likely not had the SSD long enough to really put it through its PS tests :)

I used to use short stroked discs and create 4 25 gig partitions then use those in raid 0 for my scratch and it helped massively to write large files to a large 8 disc raid 1/0 setup ! this was a while ago and I used this for a long time and was quite fast compared to a single HD
but this was a while ago ? and not on 64 bit and not on more modern gear

so the big question will be does the scratch get in the way of writing ? I still think it does :) and writing to my boot shows that I think ?

I think if you get the two 50s and use them for storage ? I might say get two 40 for the $99 special and use it as a dedicated scratch for PS $198 extra ?

I have two of the 100s RE and plan on getting two of the 40s at this price for scratch/cache files
my main files will stay on my raid though as it does quite well
if anything one of the SSD will head over to the wifes computer and I might buy her a 40 gig for scratch ! also

Honumaui

Aug 22, 2010, 12:46 AM

OH doing those tests I mentioned my scratch when writing gave me the disc full
I am sure the testing with those size files took over the 100 gig disc I had !

anyway just checked I only have 17 gigs of LR cache files on the disc so this gives you a idea that PS took over all extra 83 gigs almost ?

might help you decide how big scratch disc you are going to need ?

my PS I have to point to my boot as next in line

but the fact is if your scratch fills up you cant write files to it? that might throw a wrench in your setup you are thinking of !!!! since I hit it with a 2.8 or so gig file and only had 17 gigs of other stuff on !!!

that really surprised me as I have never seen that ! but then again I never write my files to my scratch ?

good reason not to do it in real world though :)

Giuly

Aug 22, 2010, 01:37 AM

so my raid my replicate the speed of two SSD in raid 0 Which would be 570MB/s read and 550MB/s write for the OWC SSDs we are talking about, you showed of ~375-400MB/s.

A SSD is as fast as 2.5 top-of-the-line 7200RPM drives under optimal conditions in RAID0.
If I understood you correctly, you run 8 Disks in RAID0+1? That would be 4 drives in RAID0 + Mirroring. A RAID0 of two OWC SSDs is about 120-150MB/s faster (Well, unless you have VelociRaptors or 15000RPM SAS drives in your RAID).

OP: You can install 2 2.5" drives (http://www.newegg.com/Product/Product.aspx?Item=N82E16817994072) in the optical bay (http://www.newegg.com/Product/Product.aspx?Item=N82E16817994073), you'd just need a PCIe SATA-II card (http://eshop.macsales.com/item/Sonnet%20Technology/TSATAIIE4I/). I'd rather use 4x40GB in RAID0 than 2x50GB (Or 4x25GB if OWC adds that to the Pro RE-line), because there is enough space to accommodate the drives. Waisting one SATA-II port for a slow DVD drive isn't really elegant, either. Put the drive inside an external enclosure (http://www.newegg.com/Product/ProductList.aspx?Submit=ENE&N=100007705%20600006253&IsNodeId=1&name=5.25%22) and use the port more efficiently.

Honumaui

Aug 22, 2010, 03:03 AM

meant to say might not my :)

if you read what I wrote I said I used to do raid 1/0 ? that was some time ago and posted what I use now

so you said
Which would be 570MB/s read and 550MB/s write for the OWC SSDs
this was from AJA ? or what program did you use ? or did you get these somewhere ?

using aja on the 128 file size my raid does this
write 1324.0
read 825.4
thats way faster than two raid 0 SSD ? numbers can show lots of things ? I am sure you know each system and depending on the program used can give many results being all over the place !
and I bet quickbench would give me different results ?

so why do I say they are close enough ?

MY SSD in Raid 0 using X bench and my Raid6 using X bench just the final numbers can post more but why ?
Results 457.16

my raid setup
Results 501.65

I know X bench is not so good ? but I did some tests in that when I had them set as raid 0 ? so it gives a idea compared to my raid for this comparison

so I guess since I took the time to do some PS tests for the OP with my system which I use to make a living doing PS with file sizes they work with will give the person a better real world idea

the fact I do this for a living and have SSD and have tried many setups I have very good real world experience to share ?

Giuly

Aug 22, 2010, 11:34 AM

meant to say might not my :)

if you read what I wrote I said I used to do raid 1/0 ? that was some time ago and posted what I use now

so you said

this was from AJA ? or what program did you use ? or did you get these somewhere ?

using aja on the 128 file size my raid does this
write 1324.0
read 825.4
thats way faster than two raid 0 SSD ? numbers can show lots of things ? I am sure you know each system and depending on the program used can give many results being all over the place !
and I bet quickbench would give me different results ?

so why do I say they are close enough ?

MY SSD in Raid 0 using X bench and my Raid6 using X bench just the final numbers can post more but why ?
Results 457.16

my raid setup
Results 501.65

I know X bench is not so good ? but I did some tests in that when I had them set as raid 0 ? so it gives a idea compared to my raid for this comparison

so I guess since I took the time to do some PS tests for the OP with my system which I use to make a living doing PS with file sizes they work with will give the person a better real world idea

the fact I do this for a living and have SSD and have tried many setups I have very good real world experience to share ?
The point is: I don't really understand you. Numbers show a lot, yes, if you assign them to identifiers. What does the numbers say? MB/s? MBit/s? Seconds it took to copy those files? RAID0/1 supposed to be RAID0+1, or RAID0 or RAID1? Your SSDs having a SandForce-1200 controller? Indilinx? Intel? Micron?
825.4MB/s is what 8 hard drives in RAID0 would do - or 3 of these SSDs, that's why I suggested to use a RAID0 of 4 smaller SSDs rather then 2 bigger ones.

CaoCao

Aug 22, 2010, 12:18 PM

Hi all,

I have a shiny new 6 core on the way, woot woot! After reading a few other threads and a little research, here's what I think might work for my particular workflow and the drives I can pull from my old G5. Would greatly appreciate any and all suggestions, alternatives, opinions, scathing critique, cheerful chides, etc.

Primary Use: Creative Photo Retouching & Illustration. Sometimes a single image can grow to 8 gigs or more in the working stage before the layers are simplified or flattened upon client approval.

Primary Speed Bump Incremental saves throughout the day. With large files, a save (on my G5) can take 20-30 minutes. Puts me into a catch - 22. I'm stressed that a crash will evaporate all the work I've done in the last hour, but a save might cost me 30 minutes.

Configuration & Upgrades:

24Gb RAM (3 x 8),,, 4th slot empty $1,095.99

Optical Bay : OS + Apps
60Gb OWC extreme pro SSD,,, $179.99

Bay 1 : Current Job + Scratch
First of two 50 Gig OWC SSDs in Raid 0,,, $209.99

Bay 2 : Current Job + Scratch
Second of two 50 Gig SSDs in Raid 0,,, $209.99

Bay 3 : Completed Jobs
1.5Tb 7,200 rpm HD pulled from old G5

Bay 4 : Completed Jobs
1.5Tb 7,200 rpm, HD pulled from old G5

External 1: Back Up
Existing OWC Mercury Elite 1.5 Tb Firewire

External 2 : TM Back Up
Existing OWC Mercury Elite 1.5 Tb Firewire
Small Partition for Bootable Clone of OS + Apps

Thanks in Advance! - JB
Remember, you can stuff 4x2.5" drives in the optical bay six drive Raid 0+1 sounds fun

Honumaui

Aug 22, 2010, 01:21 PM

The point is: I don't really understand you. Numbers show a lot, yes, if you assign them to identifiers. What does the numbers say? MB/s? MBit/s? Seconds it took to copy those files? RAID0/1 supposed to be RAID0+1, or RAID0 or RAID1? Your SSDs having a SandForce-1200 controller? Indilinx? Intel? Micron?

you sure seemed to understand me when you said:
Which would be 570MB/s read and 550MB/s write for the OWC SSDs we are talking about, you showed of ~375-400MB/s.

WOW so you do understand what I wrote and what SSD I have :)

how about this you post up some real world times for the OP using PS and your SSD drives rather than try to pick on someone who was helping them ?

I really think real world times with PS using OWC SSD drives will help the OP more than you worrying about a few MB/s you got off some chart ?

Honumaui

Aug 22, 2010, 02:07 PM

sure you will see this but would be a good read for you since you work with large files :)

http://macperformanceguide.com/Reviews-MacProWestmere-Photoshop-diglloydHuge.html

nanofrog

Aug 22, 2010, 03:43 PM

Primary Use: Creative Photo Retouching & Illustration. Sometimes a single image can grow to 8 gigs or more in the working stage before the layers are simplified or flattened upon client approval.

Primary Speed Bump Incremental saves throughout the day. With large files, a save (on my G5) can take 20-30 minutes. Puts me into a catch - 22. I'm stressed that a crash will evaporate all the work I've done in the last hour, but a save might cost me 30 minutes.

Configuration & Upgrades:

24Gb RAM (3 x 8),,, 4th slot empty $1,095.99

Optical Bay : OS + Apps
60Gb OWC extreme pro SSD,,, $179.99

Bay 1 : Current Job + Scratch
First of two 50 Gig OWC SSDs in Raid 0,,, $209.99

Bay 2 : Current Job + Scratch
Second of two 50 Gig SSDs in Raid 0,,, $209.99

Bay 3 : Completed Jobs
1.5Tb 7,200 rpm HD pulled from old G5

Bay 4 : Completed Jobs
1.5Tb 7,200 rpm, HD pulled from old G5

External 1: Back Up
Existing OWC Mercury Elite 1.5 Tb Firewire

External 2 : TM Back Up
Existing OWC Mercury Elite 1.5 Tb Firewire
Small Partition for Bootable Clone of OS + Apps

Thanks in Advance! - JB
Unfortunately, I wouldn't go this route.

Consumer grade SSD's (MLC based) are not meant for high write conditions, which scratch usage is (I've posted on this before, so if you're interested in the specifics, you may want to do a search on MR ;)).

You'd be better off using mechanical disks in a RAID configuration compared to SSD's. And as you're post indicates you're earning a living at it, a stripe set for working data is not a good idea either, given the time spent fixing a problem, or worse, an insufficient backup system = lost data (not just having the disks for backup, but the frequency that those backups are made; the longer the time, the more work that has to be re-done to get the missing data back).

SLC based disks would be sufficient, but the capacities are low, and they're quite expensive yet (i.e. Intel X25-E models). At those funds, you have better options (redundancy, still better write cycle conditions, and higher sustained throughputs for similar funds).

Another thing to consider is that the ICH (SATA controller) in the system has a throughput limit of ~660MB/s. You'd throttle with 3x of those SSD's (~250MB/s each, so simultaneous access will throttle).

Option 1:

Use an SSD as a OS/applications disk in the empty optical bay (boot/applications disk)
Substitute mechanical disks for the SSD's in the stripe set (better suited for scratch)
Use the other mechanical disks as the primary data location (working data, as it's safer than a stripe set)
External backups as configured

The advantage here, is lower cost (it's cheaper than your original configuration), and better safety for your working data. But it's no where near where you should be IMO for earning a living with the system (for a hobbyist, it would be acceptable, as the data's not critical).

Option 2:

Use an SSD as a OS/applications disk in the empty optical bay (boot/applications disk)
4x Mechanical disks in HDD bays 1 - 4, in a RAID 10 configuration (speed is ~ that of a 2 disk mechanical stripe set, but has the redundancy of 2x disks before data is gone). This is used for scratch and primary data (working files)
External backups/archival locations (can use single disks or JBOD; eSATA card and possibly a Port Multiplier enclosure would be less expensive over time, as you just add disks)

This is still inexpensive (especially for what you get), as you now have the minimum performance requirement, and some redundancy as well. But the performance isn't as good as it could be.

Option 3:

Use an SSD as a OS/applications disk in the empty optical bay (boot/applications disk)
Use a proper RAID card, and use either RAID 5 or 6 (can handle the write hole issue associated with parity based arrays; software implementations, such as the Highpoint 2314 are not). Specifics can be gone over if you're interested in this configuration, as there's more information and options to consider (internal, external, mixed = hybrid, disk count, future expansion requirements, OS requirements, boot requirements,...).
External backups/archival locations (can use single disks or JBOD; eSATA card and possibly a Port Multiplier enclosure would be less expensive over time, as you just add disks)

This is the best way to go IMO, and what I was referring to when I mentioned "better options for the funds" in terms of using SLC based SSD's.

You've more options and expansion capabilities with this route as well (i.e. use a hybrid = internal + external disks in the array). This is why the port count matters, disk count affects the arrays possible (5/6; or even nested parity 50/60, though I doubt you'd need to go this route). IF you've sufficient ports, you can increase capacity and performance just by adding disks (really nice, and the redundancy is a necessity given what you're doing with the system IMO).

The ARC-1222 (http://www.newegg.com/Product/Product.aspx?Item=N82E16816151039) or ARC-1222x (http://www.newegg.com/Product/Product.aspx?Item=N82E16816151066&Tpk=ARC-1222x) are good cards to take a look at, as is the ARC-1680 family (http://www.areca.com.tw/products/pcietosas1680series.htm) (12 port card+ for future expansion may be needed, depending on your capacity expansion). There's an internal adapter kit (http://www.maxupgrades.com/istore/index.cfm?fuseaction=product.display&product_id=189%20) that will allow the HDD bays to work with internal cards, and a special cable (http://www.pc-pitstop.com/sas_cables_adapters/8887-1M.asp) that can take an internal port to an external enclosure. If you're more interested in an external only solution, you need to be looking at a MiniSAS (SFF-8088) compliant enclosure (example = Stardom SohoTank ST8-U5 (http://www.stardom.com.tw/sohotank_st8_spec.html)). External cables (http://www.pc-pitstop.com/sas_cables_adapters/8888-1M.asp) per MiniSAS port (handles 4x disks).

A few notes:
With mechanical disks, you want to stay at 50% or less full for performance reasons (inner tracks = slowest on the disks, and when you get into this area, your performance can drop below the minimum requirements; particularly to be noted on a 2x disk stripe set, or even a 10 array, which only offers you half the total capacity as a trade-off for the redundancy). In the case of a 10, you'd probably be best served by using 2TB disks.

With a RAID card (i.e. Areca), you need to run enterprise grade disks if you want it to work (consumer disks are unstable, so don't do it; their recovery timings are wrong for RAID cards). As a result of potential problems, it's advisable to use the HDD Compatability List to determine what drives to use (not all RAID card makers offer these, but Areca does, and is one of the reasons I like their products - saves a lot of hassle and aggravation).

You've not mentioned the need for a Windows disk, but if you create an array under OS X, you won't be able to boot Windows off of the ICH (SATA controller on the logic board). But this is fixable via a separate SATA controller and cable assembly (you'll have to make this by splicing 2x cables that are available together). Not expensive either (card + cables), and not hard to do.

RAID wiki (http://en.wikipedia.org/wiki/RAID) might be a good thing to give a good look at, particularly 10/5/6 levels.

Also, you'd need to run a good UPS system as well (Online type is really what you should be using, though a Line interactive can be substituted so long as it has a step transformer in a budget pinch). BTW, Online types can be had refurbished to save on funds as well. A UPS isn't just an option with RAID, it's a necessity (you'll be burnt in terms of lost data if you try running without it).

I know this is a lot to read, and hope not too confusing, but it should help. :)

JulianBoolean

Aug 22, 2010, 03:53 PM

nanofrog - I was hoping you'd find my post! Wow, lots for me to sift through and absorb. Big thanks man. I'll be back with questions. -JB

nanofrog

Aug 22, 2010, 03:58 PM

nanofrog - I was hoping you'd find my post! Wow, lots for me to sift through and absorb. Big thanks man. I'll be back with questions. -JB
:cool: NP :)

Just send me an address, so I can send a bill... :eek: :D :p :p

BTW, Areca's just announced their 6.0Gb/s RAID products (http://www.areca.com.tw/news/2010/n080910.htm) (1880 series (http://www.areca.com.tw/products/1880.htm)), and ATTO's are actually shipping as I understand it. :)

Edit: The ARC-1880 series are available here (http://www.pc-pitstop.com/sas_controllers/#Areca). ATTO's 6.0Gb/s RAID versions are still hard to find (H series = non RAID HBA are available).

Honumaui

Aug 22, 2010, 04:08 PM

glad he jumped in as he kinda backs up what I said :)

and not sure why the other guy wants to get into a wacky match ?? OH well

I can say my system I make a living off is

mac pro 3,1 8 core 2.8 with 14 gigs (want more memory)

1222x card in raid 6
OWC SSD as boot and another for cache/scratch
also he backs up what I say about not using raid 0 ?

I use stand alone raid 5 boxes to backup to and to run Time Machine

one thought on the $99 40 gig OWC setups ? for me if they last a year it will pay for itself my cache is more for read and access speed coming off LR
the scratch is more when PS hits it ?

Nanofrog a question how much do you really think it takes to destroy a OWC SSD type drive for scratch ?

again curious as my thought was if it lasts a year or two ? thats all I need cause I am sure by then newer technology will be out to replace these ?

is that reasonable to think they might last up to 2 years or do you think 6 months or ?

Ryan P

Aug 22, 2010, 04:17 PM

Nano,

Thanks for that excellent writeup. I learned several things. I'm still a little murky on the RAID and Bootcamp though. Do you think you could expand a bit?

For example, if you had a RAID setup on an Areca card, could the Bootcamp partition go on the Areca RAID, or do you still need a separate SATA controller card with an attached harddrive to place the partition? Alternatively, if the only defined RAID array on your system was off the ICH could the Bootcamp partition go on a partition attached to a ICH attached harddrive.

Thanks!

Giuly

Aug 22, 2010, 04:24 PM

OP, btw: How many graphics cards are you intend to use? Because there is the OCZ RevoDrive (http://www.newegg.com/Product/ProductList.aspx?Submit=ENE&IsNodeId=1&Description=RevoDrive&bop=And&Order=PRICE&PageSize=20), which is a RAID0 of 2 SandForce-1200 drives combined on a PCIe card rather than attached via SATA, if you decide to go SSD.

WOW so you do understand what I wrote(...)
Well, I calculated it out.
and what SSD I have
Yes, but only since your last post.

and 28% Enterprise class over-provisioning VS 7% over-provisioning
You and OWC call it "Enterprise-Class", I call it last-gen SF-1200 drive (http://www.newegg.com/Product/Product.aspx?Item=N82E16820231362&cm_re=SSD_100GB-_-20-231-362-_-Product) (vs. latest-gen (http://www.newegg.com/Product/Product.aspx?Item=N82E16820231378&cm_re=phoenix-_-20-231-378-_-Product) SF-1200 drive).

nanofrog

Aug 22, 2010, 04:38 PM

glad he jumped in as he kinda backs up what I said :)

I can say my system I make a living off is

mac pro 3,1 8 core 2.8 with 14 gigs (want more memory)

1222x card in raid 6
OWC SSD as boot and another for cache/scratch
also he backs up what I say about not using raid 0 ?

I use stand alone raid 5 boxes to backup to and to run Time Machine

one thought on the $99 40 gig OWC setups ? for me if they last a year it will pay for itself my cache is more for read and access speed coming off LR
the scratch is more when PS hits it ?

Nanofrog a question how much do you really think it takes to destroy a OWC SSD type drive for scratch ?

again curious as my thought was if it lasts a year or two ? thats all I need cause I am sure by then newer technology will be out to replace these ?

is that reasonable to think they might last up to 2 years or do you think 6 months or ?
RAID 0 definitely isn't meant for those that rely on data, so basically it's not usable for pros, save scratch space (temp data).

It's fine for enthusiasts that can deal with the time involved when a failure occurs (most data is likely off of installation disks, not content they generated themselves, such as gamers needing to get the data off of the disk fast enough to load the game data needed for the graphics card).

I know there's pros that do use it, but it's either a result of insufficient budgets, or lack of understanding of RAID or their actual needs. :( I fear for such users, as they're basically playing Russian Roulette with their data.

Nano,

Thanks for that excellent writeup. I learned several things. I'm still a little murky on the RAID and Bootcamp though. Do you think you could expand a bit?

For example, if you had a RAID on an Areca card could Bootcamp go on the Areca RAID or do you still need a separate SATA controller card? Could you get away with it if you didn't create a RAID on the ICH and only on the ARECA.

Thanks!
It has to do with users that want to create a RAID via Disk Utility and run a Windows or Linux disk (all the disks are on the same controller). Disk Utility mods the system's firmware, and Windows/Linux disks will no longer boot (assuming BC was used prior to the RAID creation). If you do the RAID first, the Windows installation won't ever boot up.

By separating them to different controllers, a user can have both. This can be as simple as a SATA card for the Windows/Linux disk, or a proper RAID card for the array (though if there's another array created under Disk Utility, the separate SATA controller would also be necessary if the RAID card is set to boot EFI). IF not, and there's an available port (RAID card still boots BIOS), then you can attach another disk as a single, and use that.

Boot Camp won't work actually with a RAID card (allows both OS's on the same disk/array), but in such an instance (separate disk), you don't need it anyway, as it's a partition tool. You can actually use it (allows for clone applications that work under OS X to work), but I prefer to use Acronis myself (use it for Windows and Linux).

If for some odd reason, you leave the card as BIOS, and want to share the array between 2x OS's, that's possible too via partitions (you do use Disk Utility, but only to partition it; not to create the array). In this instance, you can have an OS X partition and bootable Windows/Linux partition (each appears as it's own array).

But I prefer to use separate disks (each gets it's own array in such a case). Better for recovery, as one array goes down, the other is still operational. This is helpful for single disk OS installs too (makes the recovery process faster, if you need to hunt down firmware, online documentation, and use a browser to access the card itself, which is the most important part, particularly in a MP, as you can't get to the firmware settings directly; not even if it's running Windows/Linux).

In your case, all all you need to do is install a separate disk on the ICH (system's SATA controller located on the backplane board in the 2009/10 systems), and install Windows (Disk Utility hasn't made any change to the system's firmware settings). :D

Honumaui

Aug 22, 2010, 05:34 PM

You and OWC call it "Enterprise-Class", I call it last-gen SF-1200 drive (http://www.newegg.com/Product/Product.aspx?Item=N82E16820231362&cm_re=SSD_100GB-_-20-231-362-_-Product) (vs. latest-gen (http://www.newegg.com/Product/Product.aspx?Item=N82E16820231378&cm_re=phoenix-_-20-231-378-_-Product) SF-1200 drive).

no I just quoted a fact you forgot they mention ?
since you seem to be hung up on facts you read off the net ? since its obvious you dont actually have any of these things we are talking about ?

and 28% Enterprise class over-provisioning VS 7% over-provisioning
in real words %28 vs %7 ?

get a life little boy !!!!

Honumaui

Aug 22, 2010, 05:40 PM

nanofrog

Aug 22, 2010, 07:11 PM

Nanofrog a question how much do you really think it takes to destroy a OWC SSD type drive for scratch ?

again curious as my thought was if it lasts a year or two ? thats all I need cause I am sure by then newer technology will be out to replace these ?

I keep hearing some say dont use them for scratch ? but that seems to be related to the older ones and not the sandforce ?

any thoughts ?
Hard to say really, as it depends on the specifics (available capacity for writes, configuration <single disk v. RAID>, file sizes written, write frequency <most important>,...).

Manufacturers base their data on empty drives (definitely not real world conditions) and manipulated statistics (toss out the worst 10% of all the cells). To complicate matters further, things like TRIM and garbage collection will matter as well.

Currently, I'd say consumer models are good for ~3yrs for a boot/applications disk (single disk mode), and SLC based disks for ~ 5yrs. RAID however, will reduce this significantly (especially parity based arrays). And I suspect most users' scratch usage will as well from the information posted here on MR (particularly the lack of TRIM under OS X, and garbage collection for some models will exacerbate this issue). BTW, RAID disables things like TRIM, which is just one of the reasons it's worse for SSD's.

Now consider the following:

SSD's are great for random access usage, but expensive for sequential access (ignoring available SATA ports, which can be added cheaper than than the cost of an SSD; simple SATA card for example)
Scratch relys on sequential throughputs more than random access as I understand about the applications used (application based, not OS).

This means that you can get fast sequential access via mechanical for less money than SSD, and is better suited for the job. :eek: Until this changes, I see SSD as unstable due to it's immaturity (MLC specifically), and expensive for this particular purpose (particularly applicable to SLC based disks).

The source of all the consternation:
Most of the MLC based Flash is specified at 10,000 write cycles, and SLC for 100,000 cycles.

Micron has released a newer version (actually out), called eMLC, which is good for 30,000 writes, and their SLC is good for 300,000 writes. Not bad, but not commonly used just yet AFAIK (rather new, and it's going to cost more). This is then improved upon by wear leveling = rotation scheme for executing writes (rotates through all available cells before the first is re-written again, assuming it's available).

This gets complicated by real world usage (some capacity may be consumed by existing data, particularly in cases where the disk is used as a single disk, such as OS/application boot location). That leaves less capacity for data that's going to change more frequently (i.e. downloads to scratch). The higher the write frequency, the faster the available cells will die. Pure and simple, and why scratch is more dangerous to current MLC based SSD's.

So the manufacturer data needs to be "taken with a grain of salt" so to speak. ;)

Future, from available data:
This will improve over time, but it's hard to say how long it will take, as NAND Flash has it's limits (will continue to push the specifications until that limit is reached in production parts). Then another Flash technology will take over, such as FeRAM (current FeRAM is capable of 10E16 writes :eek: <MLC = 1E4, SLC = 1E5, so it's a significant increase>, which is better than current enterprise HDD's before any wear leveling or other techniques).

Honumaui

Aug 22, 2010, 08:49 PM

thanks for the thoughts yeah was going to try to guess usage ? but man thats impossible ;)
I think the tough thing is they just have not been out long enough ?

so far to me it is proving to be a bit quicker in real world as a cache drive for Lightroom

using LR seeing about %50 gains in getting the sliders ready to work but this is more a seek issue I feel since its reading the cache not writing

LR has been tougher to test ? I did some by taking video and then look at the time code to determine when sliders are coming in to view
PS is easy with its timing and such

with the LR catalogs also on SSD their is a lot of small writing to the DB and some to the new preview files but its not that big a gain in real world

of course having more ram and CS5 in 64 bit is a huge bump

and of course I have been trying to say writing large files to disc I think and my testing my system the raid 6 on the areca has been faster than SSD even in raid 0

for me SSD are still my boot and cache for LR along with Scratch for PS but going to do some more testing on this ?

also those reading cache of bridge is a huge improvement with setting Bridge cache to a SSD

nanofrog

Aug 22, 2010, 09:14 PM

thanks for the thoughts yeah was going to try to guess usage ? but man thats impossible ;)
I think the tough thing is they just have not been out long enough ?
SSD's are getting better (i.e. newer controllers, TRIM, and garbage support, and the actual Flash is improving as well), but it's still not a mature technology just yet.

so far to me it is proving to be a bit quicker in real world as a cache drive for Lightroom

using LR seeing about %50 gains in getting the sliders ready to work but this is more a seek issue I feel since its reading the cache not writing
So long as the primary usage is reads, you won't have any significant issues with buring out cells (very high write frequency).

with the LR catalogs also on SSD their is a lot of small writing to the DB and some to the new preview files but its not that big a gain in real world

of course having more ram and CS5 in 64 bit is a huge bump

and of course I have been trying to say writing large files to disc I think and my testing my system the raid 6 on the areca has been faster than SSD even in raid 0

for me SSD are still my boot and cache for LR along with Scratch for PS but going to do some more testing on this ?
This could be helpful.

As per Photoshop, I'd really recommend getting the scratch space off of the SSD.

Not sure about LR, as I've no idea on the write frequency of that application. :confused:

also those reading cache of bridge is a huge improvement with setting Bridge cache to a SSD
For reads, absolutely. SSD's are brilliant solutions for random access reads (I presume it's random access, given your comments above).

deconstruct60

Aug 22, 2010, 09:37 PM

A. Could be wrong but, I don't think photoshop is ever accessing (writing) to the scratch drive and saving the current file at the same time, because saving your file, is a one thread deal.

Not so sure that is correct. If what you are saving needs data from the incremental states stored in scratch then they compete. Not at 100% but may be some low amount of reads mixed in there.

B. .... I don't think fragmentation will be an issue will at least 75 gigs of free space at all times.

Fragmentation isn't as big of an issue with SSD drives. Fragmentation leads to skipping to different parts of the drive to get data. That is absolutely the normal mode for a flash based SSD drive. (the data is stored all over the drive. It logically looks like it is continuous perhaps, but that is not the physical layout. )

C. I will get 2x the speed of an allready fast SSD for both tasks

Generally RAID was invented to get properties of disks that are more expensive using "inexpensive" drives. It isn't speed you are getting with RAID-0 flash SSD drives but reliability and useful lifetime.

A two element stripe RAID across flash SSD drives increases wear lifetime 2x (i.e., you have cut writes by 1/2 ). A SLC SSD drive is 6x - 10x better than a MLC drive. You just need to find three independent 2x factors to get a 8x increase ( 2 * 2 * 2 ). That will put you much closer to SLC wear lifetimes.

The wiggle part is if the internal disk controller is going to give you that 2 * 2 (or better ) improvement.

The OWC RE drives have about a 20% over provisioned. Not all of that is gc queue. some is checksum and some other duplication and management. It will save you from blowing out the 'available clean cell' write buffer. Not going to 100% contribute to better lifetime. Just makes living with TRIM more of a non issue (so whether it is passed through by RAID card or not doesn't matter. The OS isn't passing it either in the first place. )

Putting the working file as decreases the flash SSD drives lifetime. If trying to extend it. The don't put those files there. They too have to be written and erased.

The more data you place on a flash SSD drive the less space it has to juggle not reusing the same cells over and over again. In other words if some file squats on 10% of the drive that is 10% that can't be used to avoid wearing. If the squatting percentage goes to to 50% then that is just 40% more space that can't use to spread out the writes.

If going to write 5-10x more than "normal" then need to minimize the files that are squatting on the drive.

[quite]
As for the increased danger of data loss with Raid 0, I'm thinking that the external FW drive with time machine will save my bacon with auto saves every hour on the hour.
[/quote]

Time machine pointed at scratch drive is bad. You don't want old scratch stuff. It was temporary when you were dealing with it and it is usefulness after you are done. Time Machine is going to try to keep different versions that have no value.

I think this is being motivated because the working file is there. I that moves then there nothing to backup here. Minimally you'd need to restrict to the folder where storing working file ( which is a different one that scratch). So
/Volume/WorkingStuff

The scratch two top level folders

/Volume/WorkingStuff/PSScratch

/Volume/WorkingStuff/TheFile

Time Machine would only be aimed at "TheFile" folder. You can see that over the days move files in and out of that folder that you'll be lots of versions of what is essentially different things.

deconstruct60

Aug 22, 2010, 09:53 PM

External 2 : TM Back Up
Existing OWC Mercury Elite 1.5 Tb Firewire
Small Partition for Bootable Clone of OS + Apps

If TM is backing up the OS drive then if you want to maximize recovery you would want to split these onto different drives.

If your TM disk dies then you have a clone. If your clone dies then you have a TM backup. Both on same drive means if loose drive loose both. That is OK if have a one backup drive but appears to be two here. Might as well take advantage of the different devices.

The other factor is that TM keeps lots of duplicates. While clones can be exactly the same size for TM to be effective you need like 130-160% of the total volume(s) you are backing up. (sure your disks aren't 100% full but the duplicates are going to eat space. Especially if have larger, bulky files. )

deconstruct60

Aug 22, 2010, 10:37 PM

Slightly different configuration:

Configuration & Upgrades:

24Gb RAM (3 x 8),,, 4th slot empty $1,095.99

Optical Bay : OS + Apps
60Gb OWC extreme pro SSD,,, $179.99

Bay 1 : Scratch
First of two 50 Gig OWC SSDs in Raid 0,,, $209.99

Bay 2 : Scratch
Second of two 50 Gig SSDs in Raid 0,,, $209.99

replace these every 2.5-3 years regardless if stats say are good or not.
In a year, there will be eMLC drives which put you into the current SLC class of wear lifetimes without the 3-4x price multiple. Only need to use these two long enough for them to pay for themselves and then chuck them.

2 x 50GB works here because expect that scratch file size don't go over 30-40GB high water mark (that's about 3.75-5 times file size). If it is much higher than that, then you need to switch to spinning hard drives.

Likewise not surprising to see SSD as scratch disk problems if scratch file is >60-70% of disks capacity . Ideally want to be minimally down in the over provision allocation percentage space plus a small percentage increase. (e.g., would be 10-15 GB for the OWC RE drives. ).

Bay 3 : RAID-0 Partition 500 GB current job, RAID-1 Partiton Completed Jobs
1.5Tb 7,200 rpm HD pulled from old G5

Bay 4 : RAID-0 Partiton 500GB current job , RAID-1 Partiton Completed Jobs
1.5Tb 7,200 rpm, HD pulled from old G5

[assuming that since 1.5TB drives these are relatively new (less than 1.5 years old. Also somewhat assuming that these are not in a RAID-1 setup. More so represented old jobs and really old jobs. A fraction of the really old going to push somewhere else. ]

External 1: 1TB partition: Back Up Clone , 500M partition Bootable OS + Apps Clone
Existing OWC Mercury Elite 1.5 Tb Firewire

External 2 : TM Back Up of 1TB RAID-1 space
Existing OWC Mercury Elite 1.5 Tb Firewire

External 3: Warehouse drive for "really old" not-so-current projects.
New Firewire drive.

[ Files that haven't changed in 2-3 years you should get out of your back up rotation. They only suck up time and space. If haven't looked at files in a couple years take them "offline". When you update the Warehouse, clone it and then store that clone somewhere else as backup. ]

The obviates buying some expensive RAID card where going to start using partiy. RAID parity is a significant price increaser because at that point need UPS power backup in addition to data backup.

If going to commit to RAID card I'd do the following.

Don't use the optical drive space.

DISK 1 60 GB OS + Apps.

DISK 2 50 GB scratch RAID-0

DISK 3 50 GB scratch RAID-0

DISK 4 1TB Back-UP Clone , 500 GB OS+App Clone

You hand the TM disks and the "Current Job" stuff off the separate SATA controller. Primarily trying to avoid saturating the internal SATA controller... it only goes so fast. [ In part that is what is seeing when external RAID-5/6 goes faster than some internal over allocated internal drive configuration. ] Stuffing the OD slot and all four sleds is only going to expose that more. Likewise when doing "current job" cloning separate controllers for "from" and "to".

External

RAID 5 (or 6) the 1TB Current Space. 3 x 500-600 GB drives. (or 4 x in double parity case )

RAID 0 Current file 2 x 500GB

[ if can mix and slice three 1.5TB drives with a card then use them to keep drive count down. If RAID card only consumes whole drives then drive count goes up. Can go 3 way strip over those 500GB slices. since need 3 because of the parity issue anyway. Presumes not a big issue to put "current file and "current jobs" because don't do open saves on those at the same time. ]

TM backup drive. [ depending upon external drive enclosure budget either in it or firewire. ]

Warehouse drive(s) [ still firewire since really not accessing it all that much. ]

Honumaui

Aug 22, 2010, 11:01 PM

As per Photoshop, I'd really recommend getting the scratch space off of the SSD.

Not sure about LR, as I've no idea on the write frequency of that application. :confused:

For reads, absolutely. SSD's are brilliant solutions for random access reads (I presume it's random access, given your comments above).

yeah the PS scratch I am a bit in the not sure going to test stage ;)
might go back to some short stroked setups like I used to use ? have to do some time testing ? if I can save one hour the drive will pay for itself ? so that will be my basis of time opening and closing 40 - 80 large documents 250 megs to 500 megs going to check in PS about the efficiency and scratch size and try some real world testing and decide figure its worth trying :)

just good info for ya nanofrog the LR cache is a cache it writes the files out about half the size of a raw file so 5d files in the cache are about 6-7 megs and thats it sits till its needed and read no more writing back
the write takes place when you tell it to build previews usually on import and its all based on how big it is ? so their is some write as you ad more images and depending on the size how it flushes but really very very very low write issues

now the LR catalog has lots of writes both to a database so very small hits and larger hits of 700KB to 4 megs or so depending on preview sizes selected

this is also why I have been keeping my catalogs on my raid and it seems to offer the same performance

the cache when set in LR is set to the size you want and once it fills it just purges out older files and stays at that size unlike PS which you have no control over ?

so its easy to say get a 40 gig and tell it 30 gigs for cache

those looking at LR cache I posted a graph I did a bunch of timing tests and raid 0 or a single about .01 seconds difference ? so to me the raid 0 thing would only be used for smaller SSD to have the size you want ?

nanofrog

Aug 22, 2010, 11:25 PM

Not so sure that is correct. If what you are saving needs data from the incremental states stored in scratch then they compete. Not at 100% but may be some low amount of reads mixed in there.
SATA = Half Duplex
SAS and FC = Full Duplex

But keep in mind, the second channel in SAS and FC can't be used simultaneously for data (one for data, one for commands).

Even in the case you describe, the data should be read, processed by the system's CPU, and then the output written to the disk. If it's fast enough, the disk will at least have to wait for the platter to rotate to the correct position before the write can be executed (i.e. reads on the first pass, system processes the data, then writes on the second rotation; assumes the servo remains in the same position).

I'm just not aware of what you're describing actually occuring, though I think I see where you coming from (based on PMR head tech I presume). But the data still has to be processed by the system if I understand your point correctly.

Bay 1 : Scratch
First of two 50 Gig OWC SSDs in Raid 0,,, $209.99

Bay 2 : Scratch
Second of two 50 Gig SSDs in Raid 0,,, $209.99

replace these every 2.5-3 years regardless if stats say are good or not.
In a year, there will be eMLC drives which put you into the current SLC class of wear lifetimes without the 3-4x price multiple. Only need to use these two long enough for them to pay for themselves and then chuck them.
I understand your point (and a really good idea to recommend a short MTBR btw), but why not just use mechanical until the eMLC based SSD's actually ship?

Cheaper to get going, and can allow for prices to lower a bit as well. Makes more sense to me, as a mechanical array is sufficient for the intended usage.

At least SSD's don't have to deal with track locations though, which does mean for best performance, don't let the array get more than 50% full for mechanical platters. A definite advantage, but pricey ATM.

Just need sufficient time to pass before SSD's really hit the point they're able to replace mechanical for most things.

yeah the PS scratch I am a bit in the not sure going to test stage ;)
might go back to some short stroked setups like I used to use ? have to do some time testing ? if I can save one hour the drive will pay for itself ? so that will be my basis of time opening and closing 40 - 80 large documents 250 megs to 500 megs going to check in PS about the efficiency and scratch size and try some real world testing and decide figure its worth trying :)
Short stroking the disks is another nice little trick (use it myself for OS's and their clones).

It'll be interesting to see what you find out, as it could help others eek out as much performance from their systems without spending substantial funds. :)

just good info for ya nanofrog the LR cache is a cache it writes the files out about half the size of a raw file so 5d files in the cache are about 6-7 megs and thats it sits till its needed and read no more writing back
the write takes place when you tell it to build previews usually on import and its all based on how big it is ? so their is some write as you ad more images and depending on the size how it flushes but really very very very low write issues

now the LR catalog has lots of writes both to a database so very small hits and larger hits of 700KB to 4 megs or so depending on preview sizes selected

this is also why I have been keeping my catalogs on my raid and it seems to offer the same performance

the cache when set in LR is set to the size you want and once it fills it just purges out older files and stays at that size unlike PS which you have no control over ?

so its easy to say get a 40 gig and tell it 30 gigs for cache

those looking at LR cache I posted a graph I did a bunch of timing tests and raid 0 or a single about .01 seconds difference ? so to me the raid 0 thing would only be used for smaller SSD to have the size you want ?
The write frequency seems lower, but I'm not sure how rapidly the previews are generated (i.e. single? large batch?).

Assuming my understanding is correct, it would probably be fine on SSD's, but given the costs involved, it may not be worth it, given the differences between SSD and RAID (stripe set), unless my understanding of what you've tested and on what is way off.

Your post was a tad confusing, so I appoligize if this is making matters worse. :eek: :p

JulianBoolean

Aug 22, 2010, 11:47 PM

Unfortunately, I wouldn't go this route. Consumer grade SSD's (MLC based) are not meant for high write conditions, which scratch usage is (I've posted on this before, so if you're interested in the specifics, you may want to do a search on MR ;)).

Ok, got it. After some reading here and elsewhere, I'm dubious regarding the use of SSDs for writing scratch and working files. Thanks!

You'd be better off using mechanical disks in a RAID configuration compared to SSD's. And as you're post indicates you're earning a living at it, a stripe set for working data is not a good idea either, given the time spent fixing a problem, or worse, an insufficient backup system = lost data (not just having the disks for backup, but the frequency that those backups are made; the longer the time, the more work that has to be re-done to get the missing data back).

Agreed. Saving working files to a raid0 drive is risky buisness. I get it.

However, speed and safety are not mutually exclusive concepts for a pro retoucher. Allow me the luxury of extended explanation. I might still be willing to live with the risk of disk failure with raid0, in exchange for the abolishment of another type of risk. Specifically the risk that, because it takes so long to save my file, that I don't save as often as I should. When working on a 9gig, hundred layer image, there is increased risk of an application freeze or system crash. So ideally you'd save every 15 minutes or so right? Well not if every time you save, you are left checking your email for 30 minutes. So I end up not saving for hours at a time because I've got 3 art directors an 3 account managers showing up in the studio at 4.30 pm. That is the flavor of risk I'm primarily concerned with getting rid of. Everything else is frosting on the cake.

I can live with the dif between this or that filter taking 122 seconds vs 6oo seconds on a 12 core 2.93. All those speed charts on Mac Performance Guide make for interesting reading, but I need a system that can solve one freakin huge problem that photoshop has created for me. Saving is a single task, single thread operation. Unfortunately, throwing RAM, cores, or clock speed at the problem does not fix it.

SLC based disks would be sufficient, but the capacities are low, and they're quite expensive yet (i.e. Intel X25-E models). At those funds, you have better options (redundancy, still better write cycle conditions, and higher sustained throughputs for similar funds).

Understood, thanks! FWIW, capacity of an SSDs for either scratch or working files is not really an issue. I really only need 100 gigs max for currently working jobs.

Another thing to consider is that the ICH (SATA controller) in the system has a throughput limit of ~660MB/s. You'd throttle with 3x of those SSD's (~250MB/s each, so simultaneous access will throttle).

Not sure your comment here:
• Three SSDs in a raid0 would cause a throttle?
• 3 Individual SSDs in three sep ports, but accesses at same time causes a throttle?

Option 1:

Use an SSD as a OS/applications disk in the empty optical bay (boot/applications disk)
Substitute mechanical disks for the SSD's in the stripe set (better suited for scratch)
Use the other mechanical disks as the primary data location (working data, as it's safer than a stripe set)
External backups as configured

The advantage here, is lower cost (it's cheaper than your original configuration), and better safety for your working data. But it's no where near where you should be IMO for earning a living with the system (for a hobbyist, it would be acceptable, as the data's not critical).

This would indeed cut my 40 minute photoshop saves (via one x 7200 rpm HDD) in half. But 20 minutes is still kind of a deal breaker, I need faster writes.

Option 2:

Use an SSD as a OS/applications disk in the empty optical bay (boot/applications disk)
4x Mechanical disks in HDD bays 1 - 4, in a RAID 10 configuration (speed is ~ that of a 2 disk mechanical stripe set, but has the redundancy of 2x disks before data is gone). This is used for scratch and primary data (working files)
External backups/archival locations (can use single disks or JBOD; eSATA card and possibly a Port Multiplier enclosure would be less expensive over time, as you just add disks)

This is still inexpensive (especially for what you get), as you now have the minimum performance requirement, and some redundancy as well. But the performance isn't as good as it could be.

Awesome, I like this one a bit better as well but I still get the same write times as option A.

Option 3:

Use an SSD as a OS/applications disk in the empty optical bay (boot/applications disk)
Use a proper RAID card, and use either RAID 5 or 6 (can handle the write hole issue associated with parity based arrays; software implementations, such as the Highpoint 2314 are not). Specifics can be gone over if you're interested in this configuration, as there's more information and options to consider (internal, external, mixed = hybrid, disk count, future expansion requirements, OS requirements, boot requirements,...).
External backups/archival locations (can use single disks or JBOD; eSATA card and possibly a Port Multiplier enclosure would be less expensive over time, as you just add disks)

This is the best way to go IMO, and what I was referring to when I mentioned "better options for the funds" in terms of using SLC based SSD's.

You've piqued my curiosity on this option. I'd be interested in any solution at almost any price, in order to get the saving time (write times) of my currently working images down to a fraction of what it takes on my current system(s)

You've more options and expansion capabilities with this route as well (i.e. use a hybrid = internal + external disks in the array). This is why the port count matters, disk count affects the arrays possible (5/6; or even nested parity 50/60, though I doubt you'd need to go this route). IF you've sufficient ports, you can increase capacity and performance just by adding disks (really nice, and the redundancy is a necessity given what you're doing with the system IMO).

Got it, thanks!

The ARC-1222 (http://www.newegg.com/Product/Product.aspx?Item=N82E16816151039) or ARC-1222x (http://www.newegg.com/Product/Product.aspx?Item=N82E16816151066&Tpk=ARC-1222x) are good cards to take a look at, as is the ARC-1680 family (http://www.areca.com.tw/products/pcietosas1680series.htm) (12 port card+ for future expansion may be needed, depending on your capacity expansion). There's an internal adapter kit (http://www.maxupgrades.com/istore/index.cfm?fuseaction=product.display&product_id=189%20) that will allow the HDD bays to work with internal cards, and a special cable (http://www.pc-pitstop.com/sas_cables_adapters/8887-1M.asp) that can take an internal port to an external enclosure. If you're more interested in an external only solution, you need to be looking at a MiniSAS (SFF-8088) compliant enclosure (example = Stardom SohoTank ST8-U5 (http://www.stardom.com.tw/sohotank_st8_spec.html)). External cables (http://www.pc-pitstop.com/sas_cables_adapters/8888-1M.asp) per MiniSAS port (handles 4x disks).

Awesome links. I'm terrible at finding things I don't know I might need. Big thanks on this .

A few notes:
With mechanical disks, you want to stay at 50% or less full for performance reasons (inner tracks = slowest on the disks, and when you get into this area, your performance can drop below the minimum requirements; particularly to be noted on a 2x disk stripe set, or even a 10 array, which only offers you half the total capacity as a trade-off for the redundancy). In the case of a 10, you'd probably be best served by using 2TB disks.

Got it, understood.

With a RAID card (i.e. Areca), you need to run enterprise grade disks if you want it to work (consumer disks are unstable, so don't do it; their recovery timings are wrong for RAID cards). As a result of potential problems, it's advisable to use the HDD Compatability List to determine what drives to use (not all RAID card makers offer these, but Areca does, and is one of the reasons I like their products - saves a lot of hassle and aggravation).

Got it, thanks!

You've not mentioned the need for a Windows disk, but if you create an array under OS X, you won't be able to boot Windows off of the ICH (SATA controller on the logic board). But this is fixable via a separate SATA controller and cable assembly (you'll have to make this by splicing 2x cables that are available together). Not expensive either (card + cables), and not hard to do.

Nope don't need a windows disk.

RAID wiki (http://en.wikipedia.org/wiki/RAID) might be a good thing to give a good look at, particularly 10/5/6 levels.

I will for sure dive into that link, thanks!

Also, you'd need to run a good UPS system as well (Online type is really what you should be using, though a Line interactive can be substituted so long as it has a step transformer in a budget pinch). BTW, Online types can be had refurbished to save on funds as well. A UPS isn't just an option with RAID, it's a necessity (you'll be burnt in terms of lost data if you try running without it).

Uhmm.. the edge of my knowledge forest arrives right around what you are stating above. Sounds kinda important though.

I know this is a lot to read, and hope not too confusing, but it should help. :)

Really great stuff man. Not too confusing at all, very well articulated. I'm hitting the hay for the night. Back attcha soon.

-JB

Honumaui

Aug 22, 2010, 11:52 PM

The write frequency seems lower, but I'm not sure how rapidly the previews are generated (i.e. single? large batch?).

Assuming my understanding is correct, it would probably be fine on SSD's, but given the costs involved, it may not be worth it, given the differences between SSD and RAID (stripe set), unless my understanding of what you've tested and on what is way off.

Your post was a tad confusing, so I appoligize if this is making matters worse. :eek: :p

i write very very poor sadly :(
whats funny is for 3 years I worked for HP as a trainer ? in person in front of people I have the gift yet I cant write to save my life ! so sorry

the previews are generated in batch you import your images and use a pull down to select what size during the import they are built you can build them one at a time ? but really I dont think to many do it this way ? most do it on import
you can also select build previews as a menu item ?
it really is very low priority to write it and the write speed is not at all important
but once the data is in cache LR uses that info to generate the images you work on along with your original images !
that data is used to build the images you work on so the read of that data is critical to speed of getting to work on the images

some on normal HDs have about 1-2 seconds until you can work on a image
a short stroked hitachi 1Tb to 100 gigs takes about .78 or so seconds and the SSD takes about .53 seconds but this .25 to some who are working on thousands of images it ads up quite quickly
and moving from image to image quicker also make a nicer workflow and does not break your thoughts so actual production goes way up more than the time saved !
I have noticed about a %25 time savings
a normal job is about 5 hours now about 4 hours
so those who work in LR have to decide if that time will make up for it ?

I would say the time and cost is easy if you work with LR for a living as a photographer its worth it

if its a hobby ? raid 0 short stroked seutps are not bad

this is the timing chart I came up with over a few days of testing
and this for LR users is about how quickly the white sliders come in to view in Dev mode !

deconstruct60

Aug 23, 2010, 12:01 AM

Now consider the following:

SSD's are great for random access usage, but expensive for sequential access (ignoring available SATA ports, which can be added cheaper than than the cost of an SSD; simple SATA card for example)
Scratch relys on sequential throughputs more than random access as I understand about the applications used (application based, not OS).

This means that you can get fast sequential access via mechanical for less money than SSD, and is better suited for the job. :eek: ....

that's not particularly well supported. What is being discounted here is that the Mac OS X HFS file system can hand you sequential blocks. For very large files it can't. The automatic defrag mechanism doesn't work on large allocations.

Second, most SSD drive controllers have been tweaked so sniff out sequential reads. At that point, if there is a nice sized RAM cache on the flash disk can fill it almost as fast as it is being emptied. You will get fast rates regardless if the file system has allocated sequentail blocks or not.

You can do it with disks. Folks do it for large data warehouse DB tablespaces. However those typically take performance hits as soon as do a couple of updates to them and then try to read them again. To get fast sequential disk reads you have to line up all the blocks inside of a track and then the next blocks on exactly the next track. Once you do that for several tracks most file systems loose it. There will be a block somewhere else and you'll take a track seek latency hit. As soon as do that even 10% of the time t he SSD is going to gain ground because it is a random access problem.

As long as the flash SSD has a large enough "clean cell, ready for writing" write queue and not writing tooo much, there is not a problem.

10,000 writes means you can write to the same cell 11 times a day for 2.5 years (at 50 times a day you are at .5 years ) . If someone is spending hours processing a single photo (long enough for TM backups to trigger) how likely is it for the scratch file to be overwritten 11 times in a day ? That is being pessimistic that the drive controllers wear algorithms can't distribute it around. Can easily cut that in half if the controller has a pool of candidates that is just as big as the file.

If opening/closing 11 very large ( > 1GB) files an hour you have problems. Photoshop is going to make a copy in scratch. Then make incremental additions as make changes. That will have a bad impact on MLC SSD drive.

Honumaui

Aug 23, 2010, 12:06 AM

Julian did you read the times I had with my files ?
5.47 gig file 118 seconds to save to my raid

again I can say a raid setup might be something to look into ?

the 1222x I have with the battery module in our line of work is worth it
drop 8 WD 1TB RE3 drives in it you have a nice setup that is going to be faster for large file writes than SSD in my testing
a touch over $2000 for this setup ? but worth it for the amount and speed you get

let nanofrog jump in on this ? but the other controllers that you can extend the cache on to 4GB cache on the raid controller might help ? that I do not know

this pro PS user can not afford to loose my work again 4 hours work goes down the drain thats 4 hours to redo 4 hours you lost and that puts you 4 hours behind schedule !!!

I would really go read this also
http://macperformanceguide.com/Reviews-MacProWestmere-Photoshop-diglloydHuge.html

nanofrog

Aug 23, 2010, 12:47 AM

However, speed and safety are not mutually exclusive concepts for a pro retoucher. Allow me the luxury of extended explanation. I might still be willing to live with the risk of disk failure with raid0, in exchange for the abolishment of another type of risk. Specifically the risk that, because it takes so long to save my file, that I don't save as often as I should. When working on a 9gig, hundred layer image, there is increased risk of an application freeze or system crash. So ideally you'd save every 15 minutes or so right? Well not if every time you save, you are left checking your email for 30 minutes. So I end up not saving for hours at a time because I've got 3 art directors an 3 account managers showing up in the studio at 4.30 pm. That is the flavor of risk I'm primarily concerned with getting rid of. Everything else is frosting on the cake.
This is usually the case, but usage pattern and budgets in particular, tend to carve out the actual solution implemented.

Redundancy and performanc cost money, and the combination of the two can get downright expensive. But if you need it, you need it.

Given the dangers associated with RAID 0 and your file saving habits, I'd recommendy you stay away from it, and go with a proper RAID card instead (either 5 or 6). This wouldn't solve your risk due to file saving frequency entirely, but reduce the amount of risk due to a primary array failure.

Not sure your comment here:
• Three SSDs in a raid0 would cause a throttle?
• 3 Individual SSDs in three sep ports, but accesses at same time causes a throttle?
Yes. If they're in a stripe set (or otherwise run simultaneously), then you're trying to shove 750MB/s over a SATA controller than can only pass ~660MB/s to the processor.

This is called throttling, as you're SSD throughput has to slow down to accomodate the SATA controller (ICH in this case).

You can get around this by using a PCIe card (such as a RAID card), as you're utilizing a bus that can handle more bandwidth.

You've piqued my curiosity on this option. I'd be interested in any solution at almost any price, in order to get the saving time (write times) of my currently working images down to a fraction of what it takes on my current system(s)
Take a good look at the links you've got in terms of the Areca's, enclosures,... as you can not only get redundancy, but additional throughput as well, depending on how far you want to take it (I can actually exceed 1GB/s with my own system).

Think about what kind of throughputs would put you where you'd want to be in (work backwards from time per process if you have to), capacity requirements over time,...

And do you want a "Working Data" array + Archival storage area for completed jobs, or a single array that contains all of it?

I ask, as it's easy to make multiple arrays on one card using separate disks (you can even partition disks that appear as separate arrays, though this doesn't bode well for performance, or redundancy if one fails either).

At any rate, there's a lot of options to consider here. Mostly performance and redundancy in your case, as your capacity requirements may not be that large (you'll still be using drives in parallel for a performance increase, and it also affects capacity, though both differ between levels on a fixed number of disks).

Think about it, and get back to me. Then we'll go from there. :)

Uhmm.. the edge of my knowledge forest arrives right around what you are stating above. Sounds kinda important though.
A UPS system is a necessity when you're dealing with RAID, especially so if you go with a parity based array on a proper RAID card, such as the Areca's linked previously.

i write very very poor sadly :(
whats funny is for 3 years I worked for HP as a trainer ? in person in front of people I have the gift yet I cant write to save my life ! so sorry

the previews are generated in batch you import your images and use a pull down to select what size during the import they are built you can build them one at a time ? but really I dont think to many do it this way ? most do it on import
you can also select build previews as a menu item ?
it really is very low priority to write it and the write speed is not at all important
but once the data is in cache LR uses that info to generate the images you work on along with your original images !
that data is used to build the images you work on so the read of that data is critical to speed of getting to work on the images

some on normal HDs have about 1-2 seconds until you can work on a image
a short stroked hitachi 1Tb to 100 gigs takes about .78 or so seconds and the SSD takes about .53 seconds but this .25 to some who are working on thousands of images it ads up quite quickly
and moving from image to image quicker also make a nicer workflow and does not break your thoughts so actual production goes way up more than the time saved !
I have noticed about a %25 time savings
a normal job is about 5 hours now about 4 hours
so those who work in LR have to decide if that time will make up for it ?

I would say the time and cost is easy if you work with LR for a living as a photographer its worth it

if its a hobby ? raid 0 short stroked seutps are not bad

this is the timing chart I came up with over a few days of testing
and this for LR users is about how quickly the white sliders come in to view in Dev mode !
Nice assesment. :)

You've hit the nail on the head as the old saying goes, as it all comes down to what is your time worth, and is the added cost justifiable in that context?

For a pro, the answer is more likely that an SSD is a valid solution for the funds. To an enthusiast, not as much, unless they've more money than brains. :eek: :p

that's not particularly well supported. What is being discounted here is that the Mac OS X HFS file system can hand you sequential blocks. For very large files it can't. The automatic defrag mechanism doesn't work on large allocations.
I was talking about application specific situations, specifically for Photoshop actually (though I didn't name it). In this case, the files are large enough that the filesystem can't assist in the way you describe.

In cases of small files, then Yes, I agree with you. And in such cases, it's up to the user to determine if the advantage is justifiable for their situation (i.e. pro earning a living vs. enthusiast/hobbyist with limited budgets, so choices have more impact, as adding A, may mean B can't be had).

For me, it comes down to what's justifiable for the funds (price/performance ratio). For some, going all out is worth it, even if it means a short MTBR (i.e. swapping out SSD's in say 1 - 1.5 years). But this will be out of budget for many here on MR, from what gets posted, and why I posted what I did (aimed at the described usage).

I didn't mean to create any confusion. :o

Julian did you read the times I had with my files ?
5.47 gig file 118 seconds to save to my raid

again I can say a raid setup might be something to look into ?

the 1222x I have with the battery module in our line of work is worth it
drop 8 WD 1TB RE3 drives in it you have a nice setup that is going to be faster for large file writes than SSD in my testing
a touch over $2000 for this setup ? but worth it for the amount and speed you get

let nanofrog jump in on this ? but the other controllers that you can extend the cache on to 4GB cache on the raid controller might help ? that I do not know

this pro PS user can not afford to loose my work again 4 hours work goes down the drain thats 4 hours to redo 4 hours you lost and that puts you 4 hours behind schedule !!!

I would really go read this also
http://macperformanceguide.com/Reviews-MacProWestmere-Photoshop-diglloydHuge.html
He seems to be interested in going with such a solution.

And in with Areca's, I usually do add a larger DIMM for cache (though 2GB sticks are the "sweet spot", as performance gains from additional cache aren't linear). :D

deconstruct60

Aug 23, 2010, 01:10 AM

When working on a 9gig, hundred layer image, there is increased risk of an application freeze or system crash. So ideally you'd save every 15 minutes or so right? Well not if every time you save, you are left checking your email for 30 minutes.

9,000MB in 30 minutes ( 1,800 seconds ) is 5MB/s . That doesn't seem like a disk problem. You have some pokey disks (or really bad controller or really bad fragmentation ) if the bulk of that time is just writing. Either it is far more data than that is being read/written somewhere else, the channel is clogged, or that is a computational problem.

Before running off and spending $1,000's on some complicated SSD and/or RAID-card set up I'd buy another 1.5TB drive and put it into drive bay 2. (for scratch and current single file ). Then the other two from the G5. and just trying things out. Get some baseline performance numbers with just using 64-bit CS5 along with the 24GB of memory before dropping megabucks. There is a good chance you will be surprised after you do that.
[ your 32-bit Photoshop was killing your performance with that 8-9GB image file with constant swapping to/from scratch. ]

deconstruct60

Aug 23, 2010, 01:30 AM

A UPS system is a necessity when you're dealing with RAID, especially so if you go with a parity based array on a proper RAID card, such as the Areca's linked previously.

For RAID levels with parity it is necessary because there are multiple writes to different drives that ideally should be atomic but are not. For 0 and 1 levels not as much. Each one of the 0 stripe updates is atomic. Likewise RAID 1 are done in sequence you haven't lost data or if parallel both failed.
Those two are no better/worse than a single drive if you cut power.

I was talking about application specific situations, specifically for Photoshop actually (though I didn't name it). In this case, the files are large enough that the filesystem can't assist in the way you describe.

Unless the application is reading/writing to raw disk they ALL go through the file system.

If look at the Download benchmark here you'll see that a SSD drive comes out on top:
http://www.anandtech.com/show/3734/seagates-momentus-xt-review-finally-a-good-hybrid-hdd/6

Honumaui

Aug 23, 2010, 01:32 AM

And in with Areca's, I usually do add a larger DIMM for cache (though 2GB sticks are the "sweet spot", as performance gains from additional cache aren't linear). :D

good to know as I was not sure exactly how it would effect that :) next year the 1800 series will be in my system and this will go down to second machine figure wait a bit let the drivers mature etc.. been following them ;) and will see what is going on in storage also to make em fly

nanofrog

Aug 23, 2010, 01:38 AM

good to know as I was not sure exactly how it would effect that :) next year the 1800 series will be in my system and this will go down to second machine figure wait a bit let the drivers mature etc.. been following them ;) and will see what is going on in storage also to make em fly
I'm waiting for performance data, particularly as they're built off of Marvell RoC controllers (want to see what they're capable of; ATTO's products too, as they're using the same Marvell chips).

For RAID levels with parity it is necessary because there are multiple writes to different drives that ideally should be atomic but are not. For 0 and 1 levels not as much. Each one of the 0 stripe updates is atomic. Likewise RAID 1 are done in sequence you haven't lost data or if parallel both failed.
Those two are no better/worse than a single drive if you cut power.
I realize what you're getting at, but I'm thinking of power lost during work (processing and writting data occur during a power outage).

Re-performing the work sucks, and it can't be relied upon that the job will resume automatically (wasn't all aimed at atomicity, but real world conditions).

So I see this as one instance where such a generalization is quite applicable, though the specific reasons differ between parity and non parity based arrays.

Unless the application is reading/writing to raw disk they ALL go through the file system.
Not that they don't all go through the filesystem, but it's ability to improve matters has limits. For small files, it can certainly give an imrovement. Not as much as the size increases, to a point it's effects are harder to measure in real world terms.

Honumaui

Aug 23, 2010, 01:49 AM

I'm waiting for performance data, particularly as they're built off of Marvell RoC controllers (want to see what they're capable of; ATTO's products too, as they're using the same Marvell chips).

yup been following this
http://arecaraid.com/forum/viewtopic.php?f=7&t=14
and the hardforums ? forgot the exact forum name ?

nanofrog

Aug 23, 2010, 04:56 AM

yup been following this
http://arecaraid.com/forum/viewtopic.php?f=7&t=14
and the hardforums ? forgot the exact forum name ?
I'll give it a read. Thanks. :)

slater-k

Aug 23, 2010, 06:32 AM

I need a system that can solve one freakin huge problem that photoshop has created for me. Saving is a single task, single thread operation. Unfortunately, throwing RAM, cores, or clock speed at the problem does not fix it.

I've got a 1680x card hooked up to 7 1tb RE3's, primarily for PS and LR ... i've just timed saving a 1.9Gb multi layer file in PS and it took 2 min 37 sec (while it's AJA The interesting thing is that according to the disc activity in iStat, the disc is not actually being written to for the vast majority of the time that the process is going on, just a spike of activity here and there (at 360 Gb/s). If this is accurate, it implies that there will be a ceiling that you will hit regardless of how fast your machine / drives are (mine is MP 3,1, Octo 2.8, 16Gb RAM)

I was going down a similar route to you, and my eternal thanks to nanofrog for all his amazing help :D:cool: He is a complete star! :D

i went with a RAID6 array on an Areca 1680x, SSD for OS in optical bay, 4x 1.5Tb drives internally for back up. I understood that if you were to lose one of the drives in a JBOD the rest would go, not realising that you could get the data back via a recovery utility. Anyhow, the internals are there backing up on a schedule via chronosync, and os backup via superduper. Because i can loose 2 drives on the RAID6 and keep working, the backup schedule doesn't need to be aggressive (ie hourly).

I can't find the link, but i think it was Jeff Schewe who wrote in a forum that PS will not write to a file or write to scratch at the same time, so you're good with the two on the same drive.

If it isn't possible to save an 8 Gb file in under 10 mins (there or there abouts, and i don't know if this is true or not, though from my timings above i'd say there's a good chance) perhaps you should go at it from the point of making your system as stable as possible ... perhaps a mac mini to play music / surf etc while your main MP is on a user with just the os and photoshop, hooked up the a RAID 5 or 6 array?

Cheers

Ryan P

Aug 23, 2010, 09:36 AM

Did you look at your CPU usage during the save of this 1.9 GB file? I was surprised to find how much quicker my saves went when upgrading my processor last time. It's still the biggest bottleneck for me in Photoshop. I often put off saving because of it and then have a crash and loose work.

slater-k

Aug 23, 2010, 11:18 AM

Did you look at your CPU usage during the save of this 1.9 GB file?
After reading your question - yes!

It's pegged at about 106-110%, whereas a surface blur filter on the same file will take it up towards 800%, so i guess that's not the bottleneck?
(using PS CS5 BTW)

Ryan P

Aug 23, 2010, 11:35 AM

After reading your question - yes!

It's pegged at about 106-110%, whereas a surface blur filter on the same file will take it up towards 800%, so i guess that's not the bottleneck?
(using PS CS5 BTW)

My thinking is it is CPU bottlenecked but Adobe hasn't bothered to multithread it.

I'm still waiting on my Mac Pro. My current system is a MBP with the i7 2.66 which has a turbo of 3.33 which replaced my core 2 duo 2.4. The i7 sped things up roughly 3 fold for saving files.

Honumaui

Aug 23, 2010, 11:46 AM

I can't find the link, but i think it was Jeff Schewe who wrote in a forum that PS will not write to a file or write to scratch at the same time, so you're good with the two on the same drive.

If it isn't possible to save an 8 Gb file in under 10 mins (there or there abouts, and i don't know if this is true or not, though from my timings above i'd say there's a good chance) perhaps you should go at it from the point of making your system as stable as possible ... perhaps a mac mini to play music / surf etc while your main MP is on a user with just the os and photoshop, hooked up the a RAID 5 or 6 array?

Cheers

curious where is your scratch pointed to ? :)

the Schewe thing are you sure he did not mean they cant write so they have to go back and forth ? meaning dont put a file and scratch on the same disks if you can ? or it will get fraged and become slow ! I am willing to bet on this over the other :)
I am never the smartest man at the table :) but I usually am pretty good about only commenting when I know :) and will admit when I am wrong :)

remember PS hits the scratch a lot and keeps it optimized as you are working ? so a dedicated scratch is always always best !

slater-k

Aug 23, 2010, 01:45 PM

curious where is your scratch pointed to ? :)
top slice partition of my raid 6 array

the Schewe thing are you sure he did not mean they cant write so they have to go back and forth ? meaning dont put a file and scratch on the same disks if you can ? or it will get fraged and become slow ! I am willing to bet on this over the other :)
I am never the smartest man at the table :) but I usually am pretty good about only commenting when I know :) and will admit when I am wrong :)

remember PS hits the scratch a lot and keeps it optimized as you are working ? so a dedicated scratch is always always best !
Hi, i've finally found the link that i'd archived, but i'm getting a 404 when clicking on it - i guess that adobe have changed their forum addresses since august 2007.

Anyhow, i'm sure it's as i remember (that PS will not write a file to disc at the same time as accessing the scratch). Reasons:
The OP's problem is that he can't do anything else while PS is saving his file. Looking at iStat while my large file is saving, there is virtually no 'read' going on (a couple of bytes at the end). The time i gave above was for a newly opened file (ie has no history states). At 400Mb/s write speed, the file of 1.9Tb should only take around 5 seconds to actually write (but not generate).

All of this seems to me to point that it won't access scratch while writing the file ... i could see the argument that it will do this as it is "generating a full resolution composite" or other parts of the process, but it seems to me that the actual file (i was saving) is largely written to disc in a matter of seconds towards the end of the process.

Cheers

Honumaui

Aug 23, 2010, 03:01 PM

I hear ya on the writing at the last moment and using the scratch in theory it might work ? might be the way it is ? dont know so wont say cause that would be wrong :)

so might be true ? not doubting you :)

in reality in the past I have found it quite dif I used to make a living doing arch interior work so HUGE files
and in testing writing to scratch on same disc as files turned against me with longer times !

this I do know and what I am going off also my own times for similar file sizes

for grins what happens when you point your scratch toward a un used disc ? like a BU or something ?

also the 1.9 is that on the disk ? meaning you hit the get info or see it in finder as 1.9

I have a image that is set to 300 dpi print 36 inches wide about 8 layers ?
its 1.95 GB on disk I do not check the maximize on close ! since I am the only one opening these :)

once you hit the button takes 2-3 seconds for the saving to pop up ? but the timing is showing 16.4 (16.5 saved again as a new file new name etc.. and 16.6 a third time) for me to save that file to my raid ?
areca 1222x with battery module
both are raid cards are IOP 348 basically the same setup
I have slower HDs than you but you have one less HDD ? not sure how that balances out but very close systems
I have a 3,1 with 14 gigs my OS is SL boot in 64 using CS5 running in 64
my scratch is pointed to the SSD ? that might go back to a raid 0 short stroked setup ? still testing

but if I point my scratch to my raid the times go up to about 18-19 seconds ? not a huge difference

even a 5.5 GB image on disc is less than 2 minutes for me ?
this of course is a PSB not a PSD :)

Honumaui

Aug 23, 2010, 03:07 PM

Ryan P

Aug 23, 2010, 03:37 PM

Just to pipe in that my similar issues with slow saving have to do with files with dozens of layers. I normally start with a 25ish MB 16bit file but often end up with files in the 1-2GB range. Lately I've been tackling it by trying to make my work flow as efficient as possible. So the files are very deep but not very wide or tall.

Macinposh

Aug 23, 2010, 03:45 PM

only thing I can think of is how many layers are in that file ?

unless layers have a lot to do with the compressing and getting it ready to write ?

now this makes sense ? again have not tested but count the layers in that file and let me try to duplicate a file with that many layers ?

wonder if this will be my new learned item for the day :) try to learn something every day :)

Say.

A normal 120mb,16bit, TIFF, save with LZF, Eos-5 file with no layers might take 10-15 secs to save with CS3+MP 1.1.

Ad 10 layers (light ones,maybe 1Gb flie size?) and it is 2 mins.

Ad 20 heavy layers (with lot of data,oo,maybe 3 gigs?) 6 mins?

Ad more layers...you get the point.

While,all the time, only one..ONE core is churning out the frigging results! It is really frustrating to watch the process. Never mind what HDD setup you have...

I really feel for the gfx artist brethen here,having to work with the said 30-40 layers..

Honumaui

Aug 23, 2010, 04:01 PM

Say.

A normal 120mb,16bit, TIFF, save with LZF, Eos-5 file with no layers might take 10-15 secs to save with CS3+MP 1.1.

Ad 10 layers (light ones,maybe 1Gb flie size?) and it is 2 mins.

Ad 20 heavy layers (with lot of data,oo,maybe 3 gigs?) 6 mins?

Ad more layers...you get the point.

While,all the time, only one..ONE core is churning out the frigging results! It is really frustrating to watch the process. Never mind what HDD setup you have...

I really feel for the gfx artist brethen here,having to work with the said 30-40 layers..

YES I understand that :) not what I was asking him !!!!!
of course each layer ads to the file size though ? so its going to take longer

so what I was asking is how many layers in his file that is 1.9 gigs
my 1.95 only has about 8

if a 2 gig file has 5 layers and another 2 gig file has 30 layers ? does that make a difference ? not meaning the same file but the actual file size ?

so does that make sense :)

got to run but going to try this later :)

Macinposh

Aug 23, 2010, 04:17 PM

if a 2 gig file has 5 layers and another 2 gig file has 30 layers ? does that make a difference ? not meaning the same file but the actual file size ?

so does that make sense :)

got to run but going to try this later :)

It is all about info on those layers.
If they contain no info=no data.
Little info=Little data.
A Lot of info = Muy data.

Ryan P

Aug 23, 2010, 04:30 PM

Honumaui

Aug 23, 2010, 07:15 PM

It is all about info on those layers.
If they contain no info=no data.
Little info=Little data.
A Lot of info = Muy data.

trust me I am pretty smart at this stuff :) again yes I understand layers have more info ?

but if a file is 2 gigs and has 8 layers and one is 2 gigs and has 30 layers ? should that matter ? after all its the same size is my point

Honumaui

Aug 23, 2010, 07:30 PM

While,all the time, only one..ONE core is churning out the frigging results! It is really frustrating to watch the process. Never mind what HDD setup you have...

I really feel for the gfx artist brethen here,having to work with the said 30-40 layers..

the one thing what you write to as in raid or single HDD or SSD and how much memory and your scratch all play a roll in the speed of the file being saved :)

so had a quick test with some files and found amount of layers seem not to matter ? at least for the ones I did ?

14 layer file 1.95 gigs 16.7 seconds
61 layer file 1.98 gigs 17.4 seconds

did a few files and each 3 times to avg

curious why slater with almost the same setup as me is taking over 2 minutes ? and if it is related to the scratch in some way ? or ??

Honumaui

Aug 23, 2010, 07:32 PM

For all we know could also vary depending on things such as smart filters and blend modes as well...ridiculous that they have not done a better job with multi-threading though. This might be a good thread to cross-post on an adobe forum.

yeah its stupid they have not done more cores ???

could be some of those things but the quick tests I did seem to think its not ? but then again I will never say never ?

I would be curious if slater wanted to ftp the file somewhere and let me see what it does on my setup ?

slater-k

Aug 24, 2010, 03:18 AM

how many layers are in that file ?

again have not tested but count the layers in that file and let me try to duplicate a file with that many layers ?

about 40, with around 8 being smart objects, lots of blurs and the like

if i merge all the layers, and then duplicate the resulting single layer about 8 times till the file is a similar size to the original 1.9Gb, then the save time stays roughly the same, even though what you see of the file is just the top layer (as it's set to normal and 100%).

PS is all the while using around 103% of the cpu, ie just the one core.

Someone linked to the specs of the .doc file format somewhere on these boards recently, and the pdf went on for over 100 pages, and maybe all this activity is just making sure that the file conforms to the spec for a .psd which is bound to be pretty complex?

slater-k

Aug 24, 2010, 03:21 AM

deleted

slater-k

Aug 24, 2010, 03:25 AM

trust me I am pretty smart at this stuff :)
but if a file is 2 gigs and has 8 layers and one is 2 gigs and has 30 layers ? should that matter ? after all its the same size is my point

it seems that layer number doesn't matter - my original file with 40ish complex layers and masks saves in the same time as the merged and then duplicated to 8 layers file.

slater-k

Aug 24, 2010, 03:52 AM

curious why slater with almost the same setup as me is taking over 2 minutes ? and if it is related to the scratch in some way ? or ??

Just realised that it's because my file is in 16 bit. When i convert the file to 8 bit (and in CS5 all the layers are kept as they were before) and save, it takes less than 20 seconds, even though the file has only gone from 1.9Gb to 1.7Gb.

That's saving to my raid, if i save the original file to a slow back up drive, it takes around the same time. Saving the 8 bit file takes to the raid takes around 20 secs but saving it to the slow HDD takes around 30. During the raid save, PS takes up 98-160% of the cpu (one core = 100%) for most of the time and it's pretty similar when saving to the HDD EXCEPT when i can see activity in the slow HDD when cpu drops to around 10%. After the write is over but the save is still on-going, cpu bounces back up to 160-100.

I conclude that some parts of the save are multi thread, some parts single thread, and that while the actual write to disc is going on, the cpu is still called into play in single thread mode, and if the target disc is fast then the cpu has the chance to max out, and so becomes the limiter; if the target disc is slow, then the cpu doesn't max out and it's the disc that is the limiter ... ?

(BTW, a 1.9Gb file that's only one white layer, saves in less than 1 second)

Honumaui

Aug 24, 2010, 04:48 AM

interesting on the bit :)
all my arch stuff was 16 :) but that was a few machines ago

funny on the white doc I dont do to many of those :) hhehehhe

but yeah good puzzle had me scratching my head for sure on why so long and the bit depth is that right in front of ya thing :)
the work I do is mostly in 8 these days unless I get a big gradient then its 16 till done squash it down at end ?

my own work I print in 16 but the files are done so its a open and close

doing a file of mine 1.94 gigs 16 bit with some layers and got 90 seconds avg to the raid with SSD as scratch
doing the raid as my scratch the time was 98 avg
they were all within 1 sec so the avg was pretty much what they were

one thing I have the write enabled on my areca and also have the battery module
all the high end machines I used to work with had batteries !! so I did this to my machine just to be safer ?
curious now might have to play and test with this turned on or off later ? if that has much to do with better times ?

sorry to keep this going things like this are puzzles to me and love to figure things out and learn more

I remember working on huge files on the G5 and figure good time to go eat dinner and come back and the file MIGHT be done :)

and with that back to work !

JulianBoolean

Aug 24, 2010, 08:42 AM

Just a quick update. I'm doing some after hours (saving) testing using the four MPs in the retouching studio at work. Will post results asap.

wondering if there's a way to test a common image perhaps doubling the end product of the speed test sticky here, then save that?

Big thanks to all contributors here, no time to reply just yet but will soon!

- Julian

Honumaui

Aug 24, 2010, 10:33 AM

jwestpro

Aug 25, 2010, 08:40 PM

Please forgive me if this is old news covered long ago but why is nobody mentioning SAS internal drives (15k rpm) as an option? From what I recall, they were at the very least a lot faster access than sata but maybe not.

Or, what happened to readily available 10k and 15k rpm sata drives as the other "mechanical" drive option?

I think I understand about the total throughput via a card or controller then going into whatever drives are available in whatever arrangement. For example, the apple site shows their raid card as having "553MB/s of sequential read performance in RAID 0".

Is it the total throughput that is the issue here with the above ideas?

I am a commercial photographer since 1998 and also run into the 1-2gb file saving forever issue but that may be partially due to having held out on an older machine for so long as well as not using cs5 yet. I've been lazy but also busy so didn't spend time digging around on forums like this to learn more.

Sort of separately, but relative due to that link http://macperformanceguide.com/Reviews-MacProWestmere-Photoshop-diglloydHuge.html

It appears the main drawback to the 6 core 3.33 is the ram max? 2.93 vs 3.33 for cs5 specifically, doesn't seem like a big deal but the ram could be for certain things, especially if you want to play in video too.

I have been using an enterprise level ibm server for my primary backup plus redundant on esata drives in both east and west coast studios - just in case.

It does sound like I should rethink the internals though of the 8 core I bought early 09 as well as the 6 or 12 core I'm buying in the next day or two.

I don't see the benefit of using an ssd for the os/boot drive though as opposed to a 10 or 15k sata because the files are not ever on that drive.

Any help is appreciated but I will follow this thread to learn as much as possible.

nanofrog

Aug 25, 2010, 10:12 PM

Please forgive me if this is old news covered long ago but why is nobody mentioning SAS internal drives (15k rpm) as an option? From what I recall, they were at the very least a lot faster access than sata but maybe not.
Most users here aren't willing to pay the cost/GB necessary from what I can tell. SATA is a cheaper alternative, and for sequential access situtations, can give similar performance for less money (just use more SATA drives, and it usually provides more capacity as well).

There are advantages however, such as faster random access (not as fast as SSD, but better for high write situations), and better sequential throughputs per disk (can matter where the member count is limited to available equipment, and SATA can't meet the requirements), as well as improved recovery due to SCSI commands.

Or, what happened to readily available 10k and 15k rpm sata drives as the other "mechanical" drive option?
10k rpm are available (Velociraptor), but not 15k SATA. There was mention of WD taking the Raptor/Velociraptor series to 20k rpm, but so far, hasn't surfaced.

It appears the main drawback to the 6 core 3.33 is the ram max? 2.93 vs 3.33 for cs5 specifically, doesn't seem like a big deal but the ram could be for certain things, especially if you want to play in video too.
OWC has confirmed that their 8GB sticks will work in SP systems (Quads and Hex cores for the 2010 models), so 32GB is actually possible (though more expensive than DP systems, as those can use smaller DIMM capacities to achieve the same capacity).

It does sound like I should rethink the internals though of the 8 core I bought early 09 as well as the 6 or 12 core I'm buying in the next day or two.
I'm not sure what you're getting at here for certian, but you may want to consider a proper RAID card for your primary data (i.e. run a parity array).

I don't see the benefit of using an ssd for the os/boot drive though as opposed to a 10 or 15k sata because the files are not ever on that drive.
Random Access performance for loading libraries for example (audio, or filters in Photoshop as I understand it). SSD is faster than any mechanical media for this. It's just not an ideal media for high write situations IMO for things such as scratch space.

Honumaui

Aug 25, 2010, 10:57 PM

SSD on boot is a luxury ?
its nice but not needed
NOW when OSX swaps its very nice cause its going to be much quicker

now having my boot on SSD I wont go back !!!!

on the boot its more about getting to the spot and getting that program off and into memory and SSD blows away any mechanical disc

the same reason they are so nice in Cache situations like for Lightroom or Bridge

head over to sites like anandtech and others and look at the random reads the seeks and the I/O of SSD and you will see they are so amazing

also note not all SSD are the same as mac users I think we should stick to a few and some searches here will bring those up but the Sandforce options might be the best bets

so OWC or Vertex 2 or the Gskill phoneix pros
with those 3 choices I think for photographers having your boot on a SSD and your cache files if you use LR or Bridge are worth it ?

LR or Bridge cache are not scratch :)
the thrashing and killing of a SSD used for scratch is unknown I think ? not sure if its worth it trying to do some time testing later and decide if throwing away $100 on a 40 gig SSD is worth it ?
this is if you make your living on PS work and can say well saving 4 hours a year is worth it
so for me if I am working on 100 files in a row and I can save 30 seconds a file sit and do nothing time its going to be worth it !!!!
so far last night I did a job of 50 images and watched my waiting for PS time and tomorrow my next job going to switch back to to my regular scratch and see if I notice any sit and wait for PS time ? if I dont notice any that about answers it for me

so my old short stroked raid 0 scratch might be the better option still for my scratch ?
I wil try it back and forth a bit and decide if its worth it to test farther ? if I cant notice something chances are its not worth testing ?

I have a set action to run when opening a image to edit it lays down a burn and dodge and contrast layer I can then quickly move across and adjust to get my base image ready for more ! now its instant ? I open the file it pops in ? my old setup took 2 seconds so I went about thinking how to make it quicker !

I notice in LR the biggest complaint is waiting for the sliders in Dev mode so I did a ton of testing and found SSD and putting the cache on SSD got my sliders working in about .52 seconds vs .79 for a short stroked HD so that might not sound like much but in real sit and wait it was worth it !!! thats a real world example of SSD use that for us photographers will pay off !

so boot yes !
scratch ? at this point NO keep doing short stroked raid 0 setups of at least 2 discs ?
LR and or Bridge cache YES !
LR catalogs ? no keep em on a raid
files keep em on a raid !

StofUnited

Aug 25, 2010, 11:14 PM

OMG - my head is spinning!

This thread should provide grad school credit to those who read the whole thing and pass the test at the end. What a fantastic amount of information.

You guys really add value and make this community great.

jwestpro

Aug 25, 2010, 11:30 PM

I must say thank you to the folks in this thread. I have learned a lot today reading here and on diglloyd due to the extra links provided in this thread.

As long as it's super reliable and the image information is equally intact, I may use an ssd for my working/write drive during each project and then dump all the results onto other drives. If even the best ssd drives are a couple $100, then it's worth it to me. However, I may also do this in steps to see how good it can be in a more traditional configuration with 10k drives.

Based on what I've read, going to cs5 and on a new machine is going to make a big difference in itself.

I have often been on the first batch list with Calumet for things like the 1DsMk II & III as well as both newest canon ts lenses when they were barely available to anyone but on the computer front, I've always wanted to make my old one work as long as possible.

I don't use LR, but instead I have used Capture One Pro since 2004. It runs just fine on my macbook pro which is great because I can do my first edit and raw processing while on a flight. None of my work is batched out. All of it is 16 bit into layered image and adjustment layers archived as full psd version and 8 bit for clients. Basically, no image capture ever exists in less than 2 locations, with the initial being both cards in the camera.

I store all my raw work separately from finished work files, both redundantly in two studios via a "travel" drive as well as of course, in all of my client's offices. My bigger concern is the hundreds of projects on 4x5 trans and neg before digital.

I digress, sorry.

I am going to consider:
highest grade ssd in the optical bay for os/boot
1-4 work drives - 10k sata, 15k sas, or ssd?
maybe string together one type of the above?

Do you guys just order the basic mac with your preferred video card and then remove the 1 or 2T drive putting it into some external enclosure?

Honumaui

Aug 25, 2010, 11:52 PM

back in Rob Golbraith forum days :) I was quite active on his site

I still use C1 and enjoy it but LR has its pros cons I never had to pay for C1 as I did a big test on quality etc.. that kicked off a competition from them etc.. :) not sure if you remember when they gave away 10 copies ? to the best comparison but I was the one that got them to kick off the idea :)
not sure if thats pre 2004 ? I forget :) hehehehe

C1 on SSD is quicker how much ? no idea but it feels quicker and I get my jobs I do on C1 finished in a quicker time its noticeable so I did not test

I will say think about moving files to and from a SSD ? that alone can take away the speed advantage you gained by moving to them !!!

unless you automate it somehow I still prefer working on some form of quality raid if a HD dies I dont want to loose my time ! I get into doing stuff and BOOM its gone is a mental hit also !!

I can say a quality raid card like Areca with a battery module to be safe its our living after all ! and drives like WD RE3 1TB will be quite fast and in our line of work I dont think you will notice the difference of a 8 disc raid 6 on areca with WD vs SAS ? yes in bench marks etc.. and if this was a database kind of thing etc. SAS ? but this is about working on files that are large and a low number in most sense 1000 files is low for I/O or anything its a file at a time :) is what I mean

I can say I have worked with many setups of drives in various forms for photography and a quality raid card is the best mix of things

we built a 5 node clustered server for something else I used to be involved with 3 node for the database and 2 for processing it was for database and we used SCSI back then :) SAS was not around

again for our work SATA I still think is better for the small gains you get in performance we get huge gains in storage and really some SATA for moving images are faster than some SAS

SO !
with all that the price you would pay for some SAS and controllers and such you might give a serious look to a quality card like areca 1222x or 1680x and 8 1TB RE3 and nice case for the drives
you then have a super safe fast setup for working files you can loose a HDD and keep working

you will still need a backup plan !!! remember raid is not BU never has been never will be ;)

as buying ? I buy the base and do my own thing
newegg for lots of stuff
pc-pitstop for raid controllers
OWC or Macsales.com for memory and SSD now :)

Honumaui

Aug 25, 2010, 11:56 PM

I
Based on what I've read, going to cs5 and on a new machine is going to make a big difference in itself.

I have

I wanted to hit this on its own !!

CS5 and lots of memory and SL being all in 64 bit its amazing how much quicker some things are !!!!!

yes some are about the same so if something used to take 3 second and is now 1 ?

but I used to use this defog action to take long distance haze out of shots from Maui and the haze sucked in pics sometimes :)
but it used to be like 10 seconds and the new machines its down to .4
thats one good example !!!!

jwestpro

Aug 26, 2010, 12:45 AM

so on my very old g5 ppc dual 2.0, I just witnessed a 2 minute save of a 1.1 Gb image file to the internal 10k rpm working drive. This is improved from before since I just applied all that I could learn from diglloyd's site about scratch disk and cache and plug-ins.

I had a bunch of crap building up on the scratch disk so I dumped it and now being nearly empty and using it as the working drive too, things seem to be moving a little faster again.

It's amazing how long I've had this power pc based mac and it's still not so bad running cs3. I have heard that cs4 or 5 wouldn't "play nicely" with the old ppc os ?

Honumaui

Aug 26, 2010, 01:09 AM

CS5 is really for Intel macs ? and of course SL is inel only :)

since the two together and finally getting CS5 to recognize all that memory we have makes certain things much nicer for sure

JulianBoolean

Sep 3, 2010, 01:05 AM

HonoMaui -

My new 6 core has arrived! And I've already been working on some live jobs over the weekend. Since the major speed bump for me and many others who work on big files, is the read/write times, I created a common test file that would be fairly simple for any PS user here to recreate.

1. Open a new doc, 74.7"w x 80" hi at 300ppi and CMYK. (Swop Coated_V2 in my case) that should give you one layer, at exactly 2 gigs.

2. Filter -> Noise-> Add Noise -> Gaussian, 100%

3. Duplicate that layer 3 times to give you a total of four.

4. Make the top layer multiply, the next one down screen mode, the next one down, overlay.

5. Don't flatten. Save as a .PSB large document format. Don't try saving as PSD. It wont work. The final file should be 8.18 Gbs

My Test Results
Write : 23 Minutes.
Read : 11 Min. 20 Sec

My Mac: This is from my new 6 Core, straight out of the box, no apple upgrades. IE... 3 Gigs RAM, Snow Leopard. I put in my 7200 RPM drives from my G5, so I have one boot drive, one scratch, one data, and my TM Drive.

My Photoshop Prefs:
50 History states
Cache levels at 4
PS uses 80% RAM
500 Gig dedicated scratch
Use bigger tiles plugin
Never maximize psd and PSB compatibilty

If you could spare the time, I'd really be interested in finding out your read and write times from your RAID 6 set up. I really appreciate all the insights and real world experience you've been sharing here.

Thanks to all, this place effin rocks!

Julian

EDIT : Feb 8, 2011 Just tested my six core with 32GB of ram. I'm currently able to save the same 8.18GB file in 1:32. (One minute, thirty two seconds) This is a HUGE revelation. RAM makes a huge difference in write speeds. Not often discussed!

Honumaui

Sep 3, 2010, 02:19 AM

will do it might be a day or so till I get that extra time :) but can do it for ya

I have to catch up with that work thing

jwestpro

Sep 4, 2010, 04:02 PM

anyone have thoughts on the below config? I really just use cs5 for photography projects, process my images, layered, etc up to 1-2gb, then move all the finished work externally.

2 striped 100gb ssd - then mirrored internally onto another pair
(mounted in icey docks)

into the optical bay using the various 2x drive holders

1 200gb os/boot drive with all apps
1 2T re4 or another ssd for temporary archive from the working pairs.

this requires all external archive to be mirrored in some way but what i do is have it all on one large 2T or 4T, then also on a "transport" drive that i take to the other studio where it is manually copied onto the other large esata drives or the server.

this way the work always exists in 2 locations at all times. before i was always counting on it staying alive until backed up externally.

thus why i like the idea of internal mirroring. the only reason for stripe, which will also be new for me, is for write speed. read has been quick enough but saving my large finished work files takes forever.

why would anyone mirror their os, i guess for the same reasons but luckily i've never had a boot drive fail in 10 years....

Honumaui

Sep 4, 2010, 04:24 PM

a bit confused ?

the two raid 0 SSD ? is this your working HDD ? the one that holds your working files ?

1 2TB RE4 ? just get a black ? is this bu then to your raid 0 ssd ?

with files this big where is your scratch for PS ?
and how much memory ?
and writing to SSD will help but not that much check out what works at macperformanceguide.com for huge files :)

jwestpro

Sep 4, 2010, 04:49 PM

ok, i'm confused too. can't the scratch drive have anything on it? i will have 24gb ram so i don't anticipate needing scratch space much. i can go up to 32gb ram if that removes the need for scratch space.

i was planning to have the image files i'm working on each day on the two 100gb ssd. the scratch space would also be there, but those are also then mirrored. only mirrored of course because that's where the work exists until it's copied to the 2T and externally.

jwestpro

Sep 4, 2010, 04:56 PM

my files are currently not any larger than 5700x7500 pixels or so. this will increase when the 1dsmk4 comes out or inclusion of a digital back.

i save all my layers as a PSD file but maybe that should be a tiff with layers instead?

the psd with all image and adjustment layers ends up around 1.8 gb or less. this takes 2 minutes to open on my old g5, and scratch say "2.65gb/2.38gb"

what the heck does that mean anyway? if i have 24gb ram in the new mac pro, won't that be more than enough to never need scratch space?

jwestpro

Sep 4, 2010, 05:00 PM

the main thing slow to me is the save time and i like to save a few times during the work flow, 1st after bringing in all image layers and aligning them often manually, then 2nd after all adjustment layers/work is done, then often a 3rd time when some last little thing is adjusted.

i save, then flatten and save in 16 bit, then 8 bit and sharpen for client use. i then delete the tiffs used for the composite. all raw files are on their own esata archive only used for raw files.

Honumaui

Sep 4, 2010, 05:44 PM

for fun and info in PS go up to windows and choose info !
then with the little down triangle in the upper right of that window choose panel options
then check off scratch size and efficiency and timing

you can see what your computer is doing

the timing is just that how long what you last did took ?

efficiency is memory basically if you drop below %100 you are sending info out to scratch its basically the last thing you do not the total ? so keep a eye on it as you do things

your scratch sizes
left number memory being used total (all open docs not just the current so one at a time is better)
the right is the amount of ram available

basically its going to show you your scratch info like when its going and how much etc..

these tools as you are working on things will see how your setup is doing ;)

in your case your images are taking up more memory than is available
I bet if you do things and check your efficiency you will drop below %100 at times :)
you want that number on the left to be lower than on the right basically :)

so do a few things on a image and notice the change after doing some things ? adjustment layers etc..
this shows you how as you work your file scratch needs change

again you are working with huge files under 500 megs or so this stuff is not a big deal ? its fun to see and watch though as it gives you a better understanding of PS :)

does this make sense and help :)

I would say google
CS5 panel options efficiency and scratch size
or something ? and dont take my word but always double check things :)

also learn a bit about the purge command ? after complex actions I often put a purge in it ! (Under edit menu)

sometimes I put a before snapshot in the beginning then a purge so if I want to jump back its easy and quick :)

Honumaui

Sep 4, 2010, 05:50 PM

if i have 24gb ram in the new mac pro, won't that be more than enough to never need scratch space?

the second you open PS it allocates a small amount for scratch !! think of it as getting ready incase it needs it step !!!

I am building another new pro for our layout stuff unless I decide to take it over :)

after waiting a bit to see some real world times etc.. I chose the 3.2 4 core and am putting 24 gigs in it with the thought I will check timings and such and decide if more is needed ? but I doubt it :)

when I recomend doing the math how much time vs how many documents are open etc. .
I still say the 6 core makes sense for many one has to decide on something :)

my math is that this machine uses a layout program that then puts out layered PS files when its all done in a batch
so its not a big deal to save a few seconds as other things can be done

my machine I work on image to image so every step is more critical !!
just want to say this last part incase others read my other posts and think I am contradicting what I say when recommend some to get the 6 core :)

Honumaui

Sep 4, 2010, 05:52 PM

jwestpro

Sep 4, 2010, 06:42 PM

I do not see much benefit to ssd drives now. If I have enough ram, I won't be needing scratch disk space, or at least not a dedicated ssd for it.

Secondly, the current on board sata controller is only 3gb/s but there are hdd like the 10k rpm 600 gb WD VelociRaptor that is 6gb/s.

Some people do argue over which type of drives, ssd vs hdd are more reliable over long term and heavy use. No moving parts in SSD seems the obvious winner, but may need to be reformatted every year to remain good.

I suppose one plus of the ssd though is it's read/write performance as it becomes full but then again, does it matter, who would use small ssd for archive that is not needed very often.

Can I pair 2 ssd into a raid 0 as described on the first page of this thread and then have a 2nd pair that is redundantly a mirror of the first pair? raid 1 right? just like if you have raid 1 of 2 hdd in case one fails.

is raid 0 combining 2 ssd into one space, the same as "striping" across multiple ssd? for reference, like in diglloyd's read/write test of 2,3 and 4 ssd stripe sets.

i can then have 1 2T re4 internally which is also manually copied to an esata 2T to be safe.

nanofrog

Sep 4, 2010, 07:42 PM

I do not see much benefit to ssd drives now. If I have enough ram, I won't be needing scratch disk space, or at least not a dedicated ssd for it.
It's my understanding that PS is still more likely to go to scratch.

As it happens, you can get a mechanical stripe set put together on the cheap, but given the OWC 40GB unit with the same or better sustained throughput and only $100, it may still be a consideration as a scratch location (keep in mind the 1 - 1.5 MTBR; posted on this in the other thread you're discussing RAID possibilities).

Secondly, the current on board sata controller is only 3gb/s but there are hdd like the 10k rpm 600 gb WD VelociRaptor that is 6gb/s.
Keep in mind however, that mechanical disks can't even saturate 3.0Gb/s. Recent SSD units can (i.e. C300 claims 330MB/s sustained transfers, and SATA 3.0Gb/s tops out at 270 - 275MB/s).

The reason the mechanical disks are going to 6.0Gb/s is more to do with parts availability (semiconductors used on the drive PCB) and marketing than anything else.

Some people do argue over which type of drives, ssd vs hdd are more reliable over long term and heavy use. No moving parts in SSD seems the obvious winner, but may need to be reformatted every year to remain good.
Both are valid, as it depends on the usage pattern. Mechanical components can and do break, but so do semiconductors.

The difference is, SSD's are both more reliable and faster for random and sustained throughputs (single disk) than their mechanical counterparts. But HDD's are better suited to high write conditions, which scratch space is.

There's also the cost per GB to consider. $100 can get you an SSD, but it's only 40GB, when that can get you at least 1TB on mechanical.

It all depends on the intended use.

Can I pair 2 ssd into a raid 0 as described on the first page of this thread and then have a 2nd pair that is redundantly a mirror of the first pair? raid 1 right? just like if you have raid 1 of 2 hdd in case one fails.
Go backwards. Create the mirror first (RAID 1) of each pair. Then stripe (RAID 0) the two mirrored sets you just created (OS X will allow you to do this).

But there is a potential problem, and in your case, would see it using another SSD as a boot/applications disk (described in the other thread = ICH has a throughput limit of 660MB/s, and you'd be trying to push ~750MB/s, assuming each SSD can sustain 250MB/s; some can do more).

To solve this, you'd need a proper RAID card (fast enough to accomodate the SSD's) and an adapter kit to use the internal HDD bays. That will run you ~$450 for a SAS/SATA 3.0Gb/s unit, and $624 for a 6.0Gb/s model (both are 8 ports with a sufficient processor and bandwidth to handle current SSD's; but the latter unit can continue to be upgraded with newer, faster units in the future).

As it happens, using SSD's, you could get all of this internally.

But let's keep the RAID in one or the other threads to prevent confusion. ;)

jwestpro

Sep 4, 2010, 07:55 PM

nanofrog

Sep 4, 2010, 08:12 PM

i've been reading about so many different cards, internal using mini-sas, external using esate II/III however i have not found any that have both through the same card esata and internal sata 6gb/s but maybe that's the adapter you mentioned?

the apple website actually has a rocket raid esata 6gb/s card now, 4 port, $259 and owc only has the 4 port sonnet e4p so far or a 2 port 6gb/s by newertech

it sounds like i could set up with the 5 ssd and when i find the right card, get a little more out of that set up but it will work ok without the card for now.
Highpoint's products are mostly junk. The RR43xx series is OK, but their support is terrible, and I've a feeling that's a problem for you (still seem unfamiliar with RAID).

Nor do you want to use adapters with SATA signals, as the voltages are too low (data signals are only 600mV, where as SAS are 20V, and what the adapters are for).

The adapter I was referring to, was this (http://www.maxupgrades.com/istore/index.cfm?fuseaction=product.display&product_id=189%20) (allows you to use the HDD bays with an internal RAID or non RAID HBA that uses an internal MiniSAS port, also known as an SFF-8087).

Card Examples:

6.0Gb/s RAID card (http://www.pc-pitstop.com/sas_controllers/ARC1880I.asp)
6.0Gb/s non RAID HBA (http://www.pc-pitstop.com/sas_hba/attoH608.asp) (Host Bus Adapter)
3.0Gb/s RAID card (http://www.ewiz.com/detail.php?name=ARC1222&title=Areca-ARC-1222-8-Port-Low-Profile-PCI-E-x8-SAS-RAID-Controller)

The newertech card was meant for 1x disk per port, and if you get the correct version, can also support Port Multiplier based enclosures (allows for up to 5x disks on a single eSATA port). It could be used for SSD's in a pinch up to 2x (throughput limit of the card is 500MB/s in a Gen 2.0 slot).

JulianBoolean

Sep 5, 2010, 01:08 PM

You've more options and expansion capabilities with this route as well (i.e. use a hybrid = internal + external disks in the array). This is why the port count matters, disk count affects the arrays possible (5/6; or even nested parity 50/60, though I doubt you'd need to go this route). IF you've sufficient ports, you can increase capacity and performance just by adding disks (really nice, and the redundancy is a necessity given what you're doing with the system IMO).

The ARC-1222 (http://www.newegg.com/Product/Product.aspx?Item=N82E16816151039) or ARC-1222x (http://www.newegg.com/Product/Product.aspx?Item=N82E16816151066&Tpk=ARC-1222x) are good cards to take a look at, as is the ARC-1680 family (http://www.areca.com.tw/products/pcietosas1680series.htm) (12 port card+ for future expansion may be needed, depending on your capacity expansion). There's an internal adapter kit (http://www.maxupgrades.com/istore/index.cfm?fuseaction=product.display&product_id=189%20) that will allow the HDD bays to work with internal cards, and a special cable (http://www.pc-pitstop.com/sas_cables_adapters/8887-1M.asp) that can take an internal port to an external enclosure. If you're more interested in an external only solution, you need to be looking at a MiniSAS (SFF-8088) compliant enclosure (example = Stardom SohoTank ST8-U5 (http://www.stardom.com.tw/sohotank_st8_spec.html)). External cables (http://www.pc-pitstop.com/sas_cables_adapters/8888-1M.asp) per MiniSAS port (handles 4x disks).

NanoFrog:

I'm leaning towards your option #3 at this point. Count me in for an Areca card if needed. Question: I'm wondering what you think about the OWC Mercury Elite-AL Pro Qx2. Could this work instead of the Stardom SohoTank ST8-U5 that you linked for me? The aptly named SohoTank, seems like a big beast offering me a plethora of options and room for expansion, but (forgive me for getting all artsy on you) it's going to look downright fuglly sitting next to my 6-core. How about using the OWC Qx2 four disk array (in raid 10) in the configuration shown below?. I'm hoping you might do your typically masterful job of pointing out any speed bottle necks or configuration inefficiencies with this setup.

-----------

On 32GB Ram
Four 8GB sticks,,, $1,500

-----------

Empty Optical Bay : Boot + Apps
OWC 100GB Extreme Pro RE SSD,,, $370
Currently using 25GB, leaves 75GB free.

-----------

Bays 1 - 4 : Scratch RAID 0
4x 640GB : Western Digital Caviar Green,,, $53 Each
SATA-II HDD 64MB Cache w/3yr Warranty

-----------

External 1-4 : Working Files, RAID 10
4x 1TB : OWC Mercury Elite-AL Pro Qx2, 3yr Warranty,,, $600
Currently using 1.1TB, leaving 900GB Free
http://eshop.macsales.com/shop/hard-drives/RAID/Desktop/
http://eshop.macsales.com/item/Other%20World%20Computing/MEQX2T4.0S/

-----------

External 5-6 : Back Up All, JBOD Span
Use existing 1.5 Tb OWC External FW Drives for Back up.
I currently have 1.1TB of Data, and plan to put 25GB of OS and Applications on a 100GB SSD
No need to back up RAID 0 Scratch Disks

-----------

On Scratch SSDs: I've given up on the idea of SSDs for scratch. I did some testing and I created a 100GB scratch file on one of my big images in no time at all, about 20 history states. 400GB of Scratch using a pair of premium OWC 200GB SSDs would cost me $1360. A pair of 40GB the lower end OWC SSDs at $118 each is still enormously tempting, but sadly, not suitable for me. If I opt for 4 mechanical drives instead, it's WAY cheaper at $212 for 2.56TB . I also dig (in a very 70's muscle car kinda way) the idea of having the inside of the Mac being all engine. Just boot, apps and scratch, with all my working data and backups strapped to the roof with externals.

On 4 Scratch Disks, RAID 0 : Since I will have my scratch array separated from my working files array, there is no need to back up the 4x 650MB (2.56 TB) scratch disk. I don't know how the video editing folks use scratch space, but for my saving habits in PS, it's just a temp scratch pad, something photoshop reaches for when it runs out of available RAM or to store some of my history states as I'm working on the file. In fact, the 2.56TBs exceeds my scratch volume requirements, and I'll probably take HonoMaui's advice and partition these down for a shorter stroke, and greater speed as well.

On RAID 5 Vs 10: On Nanofrog's sugestion, I've looked at RAID levels 5, 6, and 10. Regarding RAID 5, I do like the idea of speed + redundancy with only 3 disks, and that seems like an elegant and economical fit for the 4 bays inside the box. You could do a boot from bay 1, and then Bays 2-3-4 for RAID 5. Then put your scratch and backups on externals. I've tentatively settled on RAID 10 for now. I could be wrong, but from most accounts RAID 10 will give me faster writes than 5. I understand some of the write speed performance issues (Raid 5 vs RAID 10) have to do with sequential vs non sequential data, the speed of the RAID card controller, available cache, I/O bandwidth, etc.. but sifting through all these interconnected factors falls outside my area of expertise.

On 24GB Vs 32GB RAM: It's official, the 6 core will accept 32 gigs of ram using four 8GB sticks. It's also now common knowledge that the using the 4th slot knocks the ram down to 1 channel mode. I'm going with 32GB because I'm assuming that the additional 25% gain in RAM will outweigh the 3-15% efficiency hit with single channel mode. Quote: "OWC has confirmed that the 8GB modules do work in the 4/6-core 2010 Mac Pro! According to OWC, using 3 modules shows a ~ 15% memory bandwidth gain over 4 modules, so the configurations with a * at right are the optimal ones. Whether real-world tasks are affected by this small difference remains to be tested, but in past testing I never measured more then 3% hit from using 4 modules instead of 3." --Mac Performance Guide
http://macperformanceguide.com/blog/2010/20100819_MacPro32--macpro-memory.html

JBOD / Partition Question: Could I create a 100GB partition on one of my two existing external Firewire drives (before or after they are joined) for a bootable clone of my OS + Apps drive? Then I could have Time Machine use the remaining 2.9TB to back up 1.1TB of data. That would give me a longer term answer for both incremental TM backups and future expansion.

BTW, G5 Vs. 6-Core : I've opened up and played with some of my previous jobs I did on my old G5, doing some typical retoucher type moves. The stock 6-Core with only 3Gbs RAM is WAY faster than my G5, loaded to the hilt with 8GB Ram. This is going to be really sweet.

Thanks again for your excellent suggestions! -- Julian

Honumaui

Sep 5, 2010, 04:33 PM

I have not forgoten you wanting me to do the tests :) just have to finish up a few things might try to get to it tonight :)

interesting on your 100 gig scratch :) and cut the disks in half for short stroke ?
but a good reason I always tell people to tests and try things out with the timings scratch size and efficiency themselves :) heeheh

I know for me the two 40s will be fine and what I do is just point the next scratch in line to my raid ? that way if I get over flow it moves on but I dont work on large files like you anymore :) thank goodness cause it was back in the G5 days :) SLLLOOOWWWWW
eat lunch check not done go to beach check not done eat dinner check OK next step !!!!

sounds like a sweet setup
my only thing might be the raid boxes ? and will try to kick that test out for you so you can see what a Areca 1222x can do in relation to something ;)
I still think a true raid card is going to spank those boxes big time but NanoFrog can tackle that explination :)

nanofrog

Sep 5, 2010, 05:28 PM

NanoFrog:

I'm leaning towards your option #3 at this point. Count me in for an Areca card if needed. Question: I'm wondering what you think about the OWC Mercury Elite-AL Pro Qx2. Could this work instead of the Stardom SohoTank ST8-U5 that you linked for me? The aptly named SohoTank, seems like a big beast offering me a plethora of options and room for expansion, but (forgive me for getting all artsy on you) it's going to look downright fuglly sitting next to my 6-core. How about using the OWC Qx2 four disk array (in raid 10) in the configuration shown below?. I'm hoping you might do your typically masterful job of pointing out any speed bottle necks or configuration inefficiencies with this setup.
The cards and enclosure gear listed was meant to be used for your PRIMARY working data.

The OWC Qx2 isn't that fast, and is really only suitable for backup given what you've posted.

As per external enclosures, there's other brands, and even other models from Sans Digital that would be better suited in terms of appearance I think.

Sans Digital TR8X (http://www.directron.com/tr8x.html)
Enhance Technology's EnhanceBOX series (http://www.enhance-tech.com/products/desktop/desktop-jbod.html) (black ones were made in silver as well, but they seem to be harder to find now).

There's even newer RAID cards out that could be an option for you, such as Areca's 1880 series (6.0Gb/s compliant).

But without exact details, I'm not sure how many ports you need, which will affect the card and enclosure selection, as well as whether or not you want/should use the internal HDD bays with the card. :confused: If you wish to go this route, let me know, and we'll go over it (every RAID solution is specific to the user's needs; there is no one size fits all). But it will allow you better performance and capacity usage vs. 10, and in the case of the ICH and Disk Utility, will keep the processing off of the system's CPU.

You can get past the SATA port limit in the MP by using non-RAID Host Bus Adapters, but it still requires a software implementation, and that means the system's resources (CPU, RAM) will be used to do the calculations. Not that big a deal for a 4x member set, but as it gets larger, so will the CPU cycles needed to run it. Just something to keep in mind.

You also need to realize that if you do go with a proper RAID card, the disks used will need to be enterprise models, such as RE3 or RE4 series from Western Digital (RE = RAID Edition). The reason is primarily the recovery timings, but they're also more rugged models (additional sensors and better specifications for things like MTBF and Unrecoverable Bit Error rates compared to their consumer counterparts).

Not a bad idea to use them in RAID, even if it's a software implementation such as Disk Utility.

On 32GB Ram
Four 8GB sticks,,, $1,500

Empty Optical Bay : Boot + Apps
OWC 100GB Extreme Pro RE SSD,,, $370
Currently using 25GB, leaves 75GB free.
These will be fine (nice that OWC did confirm the 8GB sticks will work). :)

Bays 1 - 4 : Scratch RAID 0
4x 640GB : Western Digital Caviar Green,,, $53 Each
SATA-II HDD 64MB Cache w/3yr Warranty

External 1-4 : Working Files, RAID 10
4x 1TB : OWC Mercury Elite-AL Pro Qx2, 3yr Warranty,,, $600
Currently using 1.1TB, leaving 900GB Free
http://eshop.macsales.com/shop/hard-drives/RAID/Desktop/
http://eshop.macsales.com/item/Other%20World%20Computing/MEQX2T4.0S/
Not sure what you're doing here, as the parity based array is meant to replace the 10 array for the primary working data.

Another important note, is that you want the scratch space to be no faster than the primary array, as it's a waste of funds and added complexity for no benefit (i.e. scratch read needs to be = primary write for ideal performance).

With a parity array, you can even place the scratch space on the array (done successfully before), as the array is fast enough to accomodate both tasks (and rugged enough to take the write cycles).

You can still leave the scratch space separate (nice actually), and even have the option of using SSD/s for it, so long as you're willing to accept a MTBR (Mean Time Between Replacement) schedule of 1 - 1.5 years (toss it, and replace with a new one). The $100 price tag on the OWC 40GB disks has made this possible (cheap and fast). With more expensive models, not so much, so this is a recent development, as other inexpensive SSD's were slow (not much different than their mechanical counterparts in terms of price).

External 5-6 : Back Up All, JBOD Span
Use existing 1.5 Tb OWC External FW Drives for Back up.
I currently have 1.1TB of Data, and plan to put 25GB of OS and Applications on a 100GB SSD
No need to back up RAID 0 Scratch Disks
This is where the OWC Qx2 will be a good thing to have (backups). And you're correct in realizing you don't need to backup scratch space.

On Scratch SSDs: I've given up on the idea of SSDs for scratch. I did some testing and I created a 100GB scratch file on one of my big images in no time at all, about 20 history states. 400GB of Scratch using a pair of premium OWC 200GB SSDs would cost me $1360. A pair of 40GB the lower end OWC SSDs at $118 each is still enormously tempting, but sadly, not suitable for me. If I opt for 4 mechanical drives instead, it's WAY cheaper at $212 for 2.56TB . I also dig (in a very 70's muscle car kinda way) the idea of having the inside of the Mac being all engine. Just boot, apps and scratch, with all my working data and backups strapped to the roof with externals.

On 4 Scratch Disks, RAID 0 : Since I will have my scratch array separated from my working files array, there is no need to back up the 4x 650MB (2.56 TB) scratch disk. I don't know how the video editing folks use scratch space, but for my saving habits in PS, it's just a temp scratch pad, something photoshop reaches for when it runs out of available RAM or to store some of my history states as I'm working on the file. In fact, the 2.56TBs exceeds my scratch volume requirements, and I'll probably take HonoMaui's advice and partition these down for a shorter stroke, and greater speed as well.
2x of the 40GB disks may still be an option (if you're willing to accept the MTBR idea of 1 year, maybe 1.5 years on the outside), and it's less complex than 4x mechanical units for about the same money and performance (probably 100MB/s faster for the SSD's than mechanical, based on 2x SSD's vs. 4x mechanical running at 100MB/s each).

BTW, last I looked, the 40GB units were $100. May have expired though, not sure.

If you do go with mechanical, the short stroke partitioning will improve performance for you (no matter the member count). Nice little trick. ;)

On RAID 5 Vs 10: On Nanofrog's sugestion, I've looked at RAID levels 5, 6, and 10. Regarding RAID 5, I do like the idea of speed + redundancy with only 3 disks, and that seems like an elegant and economical fit for the 4 bays inside the box. You could do a boot from bay 1, and then Bays 2-3-4 for RAID 5. Then put your scratch and backups on externals. I've tentatively settled on RAID 10 for now. I could be wrong, but from most accounts RAID 10 will give me faster writes than 5. I understand some of the write speed performance issues (Raid 5 vs RAID 10) have to do with sequential vs non sequential data, the speed of the RAID card controller, available cache, I/O bandwidth, etc.. but sifting through all these interconnected factors falls outside my area of expertise.
Using any of the cards previously linked, you'd be fine with RAID 5. And by using a larger member count (remember, this is what the external enclosures linked are for), can tear 10 a new one. :eek: :p Same with RAID 6, though it's slower than 5, and will depend on member count vs. 10. As you scale up in member count, it will outrun 10 as well (especially if the 10 is run on the ICH, as the most you can do is 6x members if you use both optical bays as well).

JBOD / Partition Question: Could I create a 100GB partition on one of my two existing external Firewire drives (before or after they are joined) for a bootable clone of my OS + Apps drive? Then I could have Time Machine use the remaining 2.9TB to back up 1.1TB of data. That would give me a longer term answer for both incremental TM backups and future expansion.
Yes, you can do this.

JulianBoolean

Sep 5, 2010, 06:35 PM

I have not forgoten you wanting me to do the tests :) just have to finish up a few things might try to get to it tonight :)

Ya man, no worries :). I'm in no big hurry. Get the work out. Go for a swim, and grab a beer on me. By the way, you do PS for a living and you live in Hawaii? I think you just became my personal hero! I create a fair amount of print images for the Corona Light and Corona Extra Beer brands. Perhaps I could have you art direct me when It's 20 below here in Chicago and I've lost all memory or ability to visualize what a tropical beach actually looks like.

Julian

Honumaui

Sep 5, 2010, 08:37 PM

for the test are you in 16 bit mode ?

my Cache is SSD drive single only 40 gigs left on a 100 RE version (waiting on my 3 other 40s to come in) and used my extra one now for boot ? so down to one for scratch and it shares my LR catalogs :)
history states 101
cache levels 6 tile 1024
open GL off
ram set to %70 I have 14 gigs total on this box

16 bit mode
28 minutes write but this was using my computer a lot :) copying stuff off that drive across the network etc.. and doing my other work and LR open etc.. while this was going on ? figured I might as well try this and see what happens :) this is a worst case scenario :)

6.12 read !
same for the read downloading client files 3 streams worth to the HDD :)
got two more jobs to get out tonight and 3 more tomorrow :) AHAHHHHH I hate pressure :) OH and LR opening and working :) hehehehe
so lots going on while I was doing this test :)

will get some time reboot and try it out with nothing else

my thought the read ? not sure how much quicker it will get ? some I am sure
the write was pretty good though :) for how much was going on

when the new 3.2 gets here I have the ram now 24 gigs and might throw the raid on it for fun ? can run it again but that will be in a few weeks ?

JulianBoolean

Sep 6, 2010, 05:07 PM

for the test are you in 16 bit mode ?

Big thanks for taking the time to do some testing for me, much appreciated! :) My test file was created in 8 bit mode, and saves out to exactly 8.18gb. I think 16 bit mode adds something like 66% to the file size. Do you still have the file? If I had your final save size I could still extrapolate it on a MB per second basis.Still on the subject of Save/Open speed, I've found some statements on the Lloyd Chambers site that speak to this issue in no uncertain terms.

Quote: 1 Why a fast hard drive or RAID might do nothing for open/save speed. "With single-threaded open/save, Photoshop is “CPU-bound”, meaning that its running time is a function of CPU speed. In other words, if 95% of the time is spent computing, disk I/O that takes zero time can only speed things up by 5%. Example: opening a 754MB test file took 14 seconds on a 4-way striped RAID, and 18 seconds on a single moderately fast hard drive (on my Mac Pro). The modest improvement reflects the CPU-bound single-threaded operation of Photoshop CS4. Still, 18 vs 14 seconds is a 20% reduction, pointing out the value of a fast striped RAID, something far more cost effective than (for example) a 3.2GHz machine over a 2.8GHz machine, which offers only a 14% speedup.
-- Mac Performance Guide

Quote 2: Photoshop is using only one CPU core for the save operation. "The save can run only as fast as that CPU core, so hard drive speed has little influence. The 4-way striped RAID used here needs about 2 seconds to save a 722MB file. If that time were reduced to zero, then the 48 second save would still be 46 seconds—effectively the same. If you must save as PSD or compressed TIF, then your only choice is to get faster machine eg 3.2GHz instead of 2.8GHz. That could cut down a 48 second save to about 42 seconds. Not so great for a $1400 premium. Note that hard drive speed does matter for saving uncompressed TIF files; the time was cut by 56% in this example. Unfortunately, there is no option to tell Photoshop to not compress PSD files, so if PSD format is mandatory, you’re stuck with poor performance."
-- Mac Performance Guide
http://macperformanceguide.com/OptimizingPhotoshop-Configuration.html#PhotoshopOpeningSaving

My take away from those statements: I keep looking for ways to drastically slash Open/Saving times in PS. There does seem to be room for improvement by 20% or so. But not the huge improvement I'm looking for. At the end of the day the Application itself (if saving files in a.psd format is required) continues to be the primary speed bump. It seems likely that I'll get incremental, but diminishing returns on any dollars invested beyond a 20% reduction in Open/Saving times. And it seems my intuition has already lead me down the right right path! IE. The fast 3.33 GHz clock speed on my 6 core, and a four disk array RAID 0.

Thanks Again :) -Julian

Honumaui

Sep 6, 2010, 05:57 PM

ahhh OK
yeah I saved the files :)

at least that gives you a idea of files that are in 16 bit was 12.77 gigs
so not all a waste :) hehehehe

8 bit was 8.19 gigs on disc still have the file will re run with the 8 bit for ya today :) from fresh reboot

Honumaui

Sep 6, 2010, 05:59 PM

part of why I wanted to do hardware raid was in the past when I had it I loved how I could do other things and not be choked to a total stop
and could do other things yes a slow down but not total choke kinda slow down that happens with regular HDDs :)

JulianBoolean

Sep 6, 2010, 10:30 PM

I've got a 1680x card hooked up to 7 1tb RE3's, primarily for PS and LR ... i've just timed saving a 1.9Gb multi layer file in PS and it took 2 min 37 sec (while it's AJA The interesting thing is that according to the disc activity in iStat, the disc is not actually being written to for the vast majority of the time that the process is going on, just a spike of activity here and there (at 360 Gb/s). If this is accurate, it implies that there will be a ceiling that you will hit regardless of how fast your machine / drives are (mine is MP 3,1, Octo 2.8, 16Gb RAM)

I was going down a similar route to you, and my eternal thanks to nanofrog for all his amazing help :D:cool: He is a complete star! :D

i went with a RAID6 array on an Areca 1680x, SSD for OS in optical bay, 4x 1.5Tb drives internally for back up....continued.

If it isn't possible to save an 8 Gb file in under 10 mins (there or there abouts, and i don't know if this is true or not, though from my timings above i'd say there's a good chance) perhaps you should go at it from the point of making your system as stable as possible ... perhaps a mac mini to play music / surf etc while your main MP is on a user with just the os and photoshop, hooked up the a RAID 5 or 6 array?

Cheers

Slater - Sounds like really sweet setup! :)

I'm not sure I have it pictured accurately though, especially the 7x 1TB drives?
Q: Is that a four disk array for raid 6 for your working files and the 3 left over for scratch? or you have your scratch and working files on the same volume?

Wow, you saved a 1.9GB file in 2min 37 seconds. Awesome! It's so fast I'm scratching my head a bit though. This would indicate saving an 8GB file in 8-10 Minutes, way faster than the 23 it's taking me.
Q: is the 1.9GB the amount on the disk after the file is saved, or the amount of scratch, or the amount before the save?
Q Are you saving the file as a layered psd, a flattend psd, a layered tif, or a flattened tif.?

Thanks! - Julian

JulianBoolean

Sep 6, 2010, 10:34 PM

OK, I feel like I'm making some progress on this! Getting really close to Nanofrogs option #3 , Just need a little help with the details.

32GB Ram
Four 8GB sticks,,, $1,500

-------

Empty Optical Bay : OS + Apps
50GB OWC Extreme Pro RE SSD,,, $200
Currently using 20GB, leaving 30GB free

-------

Internal Bays 1-4 : RAID 0 Scratch

4x 500GB Western Digital WD RE3 (16MB Cache) ,,, $95.00 each

- or -

4x 1TB Western Digital RE3 (32 MB Cache) ,,, $150 each

-------

8 Bay External Box, Sans Digital TowerRAID ,,, $390
with an Areca Raid Card, need help choosing appropriate model from amongst the many already linked.

-------

External Bays : Working Files

6 Drives on RAID 10
- or -
5 Drives on RAID 6
-or-
some other way to utilize the eight bays?

-------

External FW : Back Up

My existing 2x 1.5TB OWC FW drives, Spanned Together for 3GB Time Machine. Plus a small 50 GB Partition for bootable clone via SuperDuper.

-------

On WD RE3's : I don't need the volume of the more expensive one, but it has 2x the cache.
is it worth the extra $50 Each?

On RAID 5,6 10 : My inner Warren Buffet is telling me never invest in anything you don't understand. RAID 10 makes total sense to me. I can't fully wrap my mind around RAID 5 or 6 so there is a bit of hesitation. I'm willing to experiment a bit when I get the box and the drives.

Thoughts, Counterpoints?

nanofrog

Sep 7, 2010, 12:56 AM

32GB Ram
Four 8GB sticks,,, $1,500

Empty Optical Bay : OS + Apps
50GB OWC Extreme Pro RE SSD,,, $200
Currently using 20GB, leaving 30GB free
As before, these are fine. ;)

Internal Bays 1-4 : RAID 0 Scratch

4x 500GB Western Digital WD RE3 (16MB Cache) ,,, $95.00 each

- or -

4x 1TB Western Digital RE3 (32 MB Cache) ,,, $150 each
You don't need a lot of capacity for scratch space, so you can chose to use smaller disks.

Another option, is to use 2x of the OWC 40GB SSD's (assuming you're willing to deal with a MTBR of 1 - 1.5 years). Another member is already using them in a stripe set, and the sustained throughput per disk is ~175MB/s IIRC (makes sense as well, as the 285MB/s throughput stated is likely burst speed, not sustained). So a pair should produce ~ 350MB/s, which is what you'd get out of a 4x disk mechanical set as well (larger drives are a bit faster than smaller ones). The advantage is less initial cost ($200 vs. $380+ for the models you're considering above), and less latency. The compromise is the MTBR. Even with mechanical, the typical MTBR is 3 years anyway, so it works out to about the same funds.

This should also be fine with the system's ICH, as ~660MB/s - 350MB/s = ~310MB/s for the OS/applications disk, which will be sufficient (SATA 3.0Gb/s is only good for ~270 - 275MB/s per disk anyway; the ~660MB/s limitation has to do with the DMI bandwidth allowed to the SATA controller in the ICH so the USB and Ethernet controllers aren't stalled <allows all 3 controllers to be used simultaneously>).

As per internal drive space, that's not a problem as you're taking the Areca's members and backups external.

So an SSD (2x members) or mechanical set (4x members) is possible with what you're doing. It's up to you which way you want to go.

8 Bay External Box, Sans Digital TowerRAID ,,, $390
with an Areca Raid Card, need help choosing appropriate model from amongst the many already linked.
So you like the look of that one I see.... :D

As per the card, do you want more ports (relevant to future expansion/performance)?
Are you going to go with SSD's in the near future (i.e. replace mechanical with SSD's on that card in say the next 5 years; partly gets to the 3.0Gb/s vs. 6.0Gb/s bit)?
What kind of budget?

External Bays : Working Files

6 Drives on RAID 10
- or -
5 Drives on RAID 6
-or-
some other way to utilize the eight bays?
Not sure what you're wanting to do here....

The card opens up a lot of options, and depending on the card, you can go up to 24 disks without the need for SAS expanders (Areca's SAS cards can actually operate up to 128 disks when using these :eek:).

So until the card model and initial member count are nailed down (goes to the port count and enclosure selection), it's too hard to determine what to go with just yet.

But if that's the enclosure you go with (assuming there's not another enclosure used), you can use up to 8x disks in any configuration the card is capable of (0/1/10/5/6/50/60 as the models you'll be considering can do all of these). Bolded = better for what you're trying to do.

External FW : Back Up

My existing 2x 1.5TB OWC FW drives, Spanned Together for 3GB Time Machine. Plus a small 50 GB Partition for bootable clone via SuperDuper.
I'm not sure of what you'll end up with yet, but I'd say this will be too little capacity.

Once we get the card, drives and enclosure/s set, then we can cover this (make sure you've sufficient capacity and performance to handle the backup in a reasonable amount of time).

There's products out there that could be of use to you if what you have won't be sufficient for all of it. They're still helpful though. ;) Clones/secondary backup locations at the very least.

On WD RE3's : I don't need the volume of the more expensive one, but it has 2x the cache.
is it worth the extra $50 Each?
Not really, as the card and member count will certainly get you some serious performance. :D

B]On RAID 5,6 10[/B] : My inner Warren Buffet is telling me never invest in anything you don't understand. RAID 10 makes total sense to me. I can't fully wrap my mind around RAID 5 or 6 so there is a bit of hesitation. I'm willing to experiment a bit when I get the box and the drives.
Each of these levels are a compromise of performance and redundancy (affects usable capacity).

RAID 10:
Failure count = 2
Performance = (n * single disk performance)/2 <n = member count>
Capacity = half of the total capacity of the set (so: (8 disks * 1TB disks)/2 = 4TB usable capacity)

RAID 5:
Failure count = 1
Capacity = (n -1) * capacity of a single disk [so: 8 disks @ 1TB = (8 -1) * 1TB = 7TB usable capacity]
Performance = ~ (n * performance of a single disk) * .85 <on average> [so: 8 * 100MB/s *.85 = 680MB/s]

RAID 6:
Failure count = 2
Capacity = (n - 2) * capacity of a single disk [(8 - 2) * 1TB = 6TB usable capacity]
Performance = ~ (n * performance of a single disk) * .75 <on average> [so: 8 * 100MB/s * .75 = 600MB/s]

Real performance data varies from card to card, and setup to setup. So the performance figures I've listed are what I've approximated from test setups (picked the 1222 and 1680 series as a reference point in terms of performance with SATA drives <WD RE3 series to be more specific>). Cache values, the card's processor speeds, and exact disk models all make a difference on the specifics (i.e. the available data on the newer 1880 series is faster than the 1222 or 1680 series cards), but hopefully it will illustrate the capacity/performance/redundancy trade-offs between the levels.

For further information on the redundancy and capacity aspects, take a look at the RAID wiki (http://en.wikipedia.org/wiki/RAID) (if you haven't already, and there's others out there too if you search). Performance data is harder to find, and usually is a result of searching out a specific model number for a card. The review will describe the test setup they used, and present the test data. General formulas are impossible to find, unlike levels 0 or 10 (even harder to find level 1 data, as there's differences between hardware and software implementations).

But above all, you must understand the card you'll be using is capable of all 3 of these, and is a proper hardware controller = it can properly handle parity based arrays (5/6 and nested parity 50/60). The cost of the equipment is the same, so you have options as to what you can do, depending on the performance, capacity, and redundancy requirments that best fit your needs.

JulianBoolean

Sep 7, 2010, 10:37 AM

Nanofrog, thanks for your typically excellent critique :) I'm giving your reply careful consideration! -- Julian

JulianBoolean

Sep 7, 2010, 10:40 AM

OP: You can install 2 2.5" drives (http://www.newegg.com/Product/Product.aspx?Item=N82E16817994072) in the optical bay (http://www.newegg.com/Product/Product.aspx?Item=N82E16817994073), you'd just need a PCIe SATA-II card (http://eshop.macsales.com/item/Sonnet%20Technology/TSATAIIE4I/). I'd rather use 4x40GB in RAID0 than 2x50GB (Or 4x25GB if OWC adds that to the Pro RE-line), because there is enough space to accommodate the drives. Waisting one SATA-II port for a slow DVD drive isn't really elegant, either. Put the drive inside an external enclosure (http://www.newegg.com/Product/ProductList.aspx?Submit=ENE&N=100007705%20600006253&IsNodeId=1&name=5.25%22) and use the port more efficiently.

I'm reconsidering the idea of SSDs for scratch. I've realized that my previous testing for scratch volume needs was based on the meager 3GB of RAM I currently have, not the 32GB of RAM I will acquire. So... I'm wondering if there is a way to put 2 SSDs in the empty optical bay? I could be reading this wrong, but the links posted above by Giuly appear to only work on G5? I called OWC and was informed that while the physical space allows for 2 SSDs in one optical bay there is only one connection in there.

Remember, you can stuff 4x2.5" drives in the optical bay six drive Raid 0+1 sounds fun

I'm assuming CaoCao is correct about the physical space available in there, but is there a way to connect 4 in one optical bay? And what about heat ?

Honumaui

Sep 7, 2010, 01:06 PM

OK fresh morning reboot and did your tests with 8 bit mode like you had :) the file is 8.19 gig as it sits on disc

the open was 64.5 seconds
the Efficiency was %91
scratch was 14.8/9.06

now the above tells me well I am %91 efficient :) why this box does not have enough ram ! (I have 14 gigs and this was a fresh reboot) but the file needs 14.8 but I only had 9.06 available for the file so that is why not %100 efficient
this amount of memory is also the last values thats why I say check both
in theory the difference of 14.8 to 9.06 is not %9 but things are never pure linear with PS it goes up and down and moves all over with memory useage
here my scratch helped me out a lot I think ? could be wrong but I was using it for sure :)
SO the scratch numbers if I had had more memory like 24 I might not have hit the scratch ! it would still be their ! but not used and your 32 should be good :)

36.85 on my SSD left quit PS it jumps to 47.31 was trying to watch how much went to scratch and the scratch numbers above seem about correct
so figure thats what PS set aside for this task
usually when it opens it sets aside 250 megs or so ?
this last part does not mean to much but I did not use tons of my scratch for these things which kinda shows at least 40 gigs IS OK sometimes :) but have a second drive checked !!!!
now remember some tasks need a lot of memory to run and that means a lot of scratch and you might need 100 gigs of scratch ? this is a unknown thing ! and everyone and every file will be different its hard to say

the save time was 1:26 one minute 26 seconds :)

Honumaui

Sep 7, 2010, 01:19 PM

I'm reconsidering the idea of SSDs for scratch. I've realized that my previous testing for scratch volume needs was based on the meager 3GB of RAM I currently have, not the 32GB of RAM I will acquire. So... I'm wondering if there is a way to put 2 SSDs in the empty optical bay? I could be reading this wrong, but the links posted above by Giuly appear to only work on G5? I called OWC and was informed that while the physical space allows for 2 SSDs in one optical bay there is only one connection in there.

I'm assuming CaoCao is correct about the physical space available in there, but is there a way to connect 4 in one optical bay? And what about heat ?

remember 3 SSD you are throttled !
so that means you wont be using your other SATA internal for anything more than two unless you do not mind the hit ?
I would say put 2 off your internal connections ? maybe one off your extra SATA in the optical and one on a icy dock sled inside and then depends on your other internals ?

OR I think I figured out my new setup ?
1 off your extra optical this will be your boot SSD
4 discs in raid 1/0 or 1+0 :) inside for safe BU and fast enough if I need to work off I can this will be 3TB discs so 6 TB working space
then a card in slot 3 and 4 that will have two SSD for scratch one off each since the card will hit 500 as the two drives might exceed that ? not sure going to have to test

you could do a raid card though like me but still do the scratch off external sata card you can run the cables up inside to your optical and just use a Y to get more power to them

I am going to put in two PM cards from OWC and see how it works running my BU off the extra port
so will be

two cards one SSD off each card and one BU box off each card ! my BU never runs when I am in PS ? so they wont be battling for bandwidth
we shall see :)
the bus is shared by 3/4 so that is my limit factor but two SSD should be fine ? in theory the two cards should give them plenty breathing room since each will be on its own card

the only other thing is take out your optical and put it on a FW case ? then you have two connections up stairs in the optical but remember you dont have much bandwidth left over all !!! really room for one more HDD ?
unless you use all internals for BU ? but then where is your boot going to go :)
see the problem I think the raid 0 scratch on SSD has to be off a external card

so a single scratch and single boot will be OK and leave some room for internal discs used as BU but not data in a raid 1/0 1+0 as you are going to limit yourself I think :)

nanofrog

Sep 7, 2010, 04:03 PM

Nanofrog, thanks for your typically excellent critique :) I'm giving your reply careful consideration! -- Julian
RAID isn't a simple thing to deal with in reality, and it's a lot of information to digest. ;)

So... I'm wondering if there is a way to put 2 SSDs in the empty optical bay? I could be reading this wrong, but the links posted above by Giuly appear to only work on G5? I called OWC and was informed that while the physical space allows for 2 SSDs in one optical bay there is only one connection in there.
Physically, it's doable.

But the ports between the two systems are different. So getting them attached to the ICH would be impossible in that location without an adapter offered by MaxUpgrades (also need a mounting system of some sort). There are other options that would be cheaper, such as pull the optical disk in ODD bay 1 (one SSD per optical bay for the cables), use HDD bays with an adapter, or use a separate controller (inexpensive unit per se, but this can get pricey too; unit I'm thinking of would be $130).

Of these, the HDD bay + adapter (here (http://www.newegg.com/Product/Product.aspx?Item=N82E16817994064&cm_re=icy_dock_2.5_3.5-_-17-994-064-_-Product); one per bay) would be the most effective IMO. Probably the cheapest too.

SSD's don't make heat like mechanical disks can, so it's nothing to worry about.

remember 3 SSD you are throttled !
so that means you wont be using your other SATA internal for anything more than two unless you do not mind the hit ?
As the 40GB units are generating about 175MB/s sustained, it would be OK to run a pair for scratch, and a separate SSD for OS/applications (leaves ~310MB/s on the ICH, and a single SATA 3.0Gb/s port tops out at 270 - 275MB/s anyway, which is fine for the model being considered).

I would say put 2 off your internal connections ? maybe one off your extra SATA in the optical and one on a icy dock sled inside and then depends on your other internals ?
There'd still be a need for 2x Icy Dock adapters (2x SSD's in the HDD bays, one in the empty optical bay) to get them installed for the lowest cost (all 3x units on the ICH).

OR I think I figured out my new setup ?
1 off your extra optical this will be your boot SSD
4 discs in raid 1/0 or 1+0 :) inside for safe BU and fast enough if I need to work off I can this will be 3TB discs so 6 TB working space
then a card in slot 3 and 4 that will have two SSD for scratch one off each since the card will hit 500 as the two drives might exceed that ? not sure going to have to test
This will work, as you're willing to put the funds into such large capacity disks and accept the trade-off in usable capacity.

I presume the RAID card is handling your primary data.... ;) :p

you could do a raid card though like me but still do the scratch off external sata card you can run the cables up inside to your optical and just use a Y to get more power to them
Let's wait and see what happens, as it may be more effective to run backups externally (JBOD, hardware RAID box via FW or even eSATA,...).

But the newertech card + pair of single disk enclosures would suffice for the scratch stripe set without getting crazy in terms of cost I think (figuring card + 2x eSATA enclosures would be under $100).

two cards one SSD off each card and one BU box off each card ! my BU never runs when I am in PS ? so they wont be battling for bandwidth
we shall see :)
You don't need to do this. :D One card will suffice (500MB/s capable, and the pair will only generate ~350MB/s, assuming you're sticking with the 40GB units from OWC).

the only other thing is take out your optical and put it on a FW case ?
USB is better, as it would work for any OS (Windows has dropped FW support all together). May not matter now, but could in the long term (i.e. decide to install a Windows installation at a later date). Bandwidth wise, USB 2.0 is fine for optical disks.

JulianBoolean

Sep 7, 2010, 04:46 PM

OK fresh morning reboot and did your tests with 8 bit mode like you had :) the file is 8.19 gig as it sits on disc

the open was 64.5 seconds
the Efficiency was %91
scratch was 14.8/9.06

the save time was 1:26 one minute 26 seconds :)

Jumpin Jahosafats Batman!!! Saving an 8GB file in 1:26! That is SSSSmokin. This is a real game changer. Your configuration, as described a few pages back is : Two 100 gig OWC RE SSD drives one for cache one for boot. Then an Areca 1222x with 8 750 drives setup raid 6 for working files and two stand along raid 5 boxes one for BackUp one for time machine and other offline offsite setups.

Question : So If I understand this correctly, what you are saving your file to is an 8 member array in RAID 6. That would be 6 disks plus 2 for parity?

Thanks again for running that test file!

JB

nanofrog

Sep 7, 2010, 04:50 PM

Jumpin Jahosafats Batman!!! Saving an 8GB file in 1:26! That is SSSSmokin. This is a real game changer. Your configuration, as described a few pages back is : Two 100 gig OWC RE SSD drives one for cache one for boot. Then an Areca 1222x with 8 750 drives setup raid 6 for working files and two stand along raid 5 boxes one for BackUp one for time machine and other offline offsite setups.

Question : So If I understand this correctly, what you are saving your file to is an 8 member array in RAID 6. That would be 6 disks plus 2 for parity?

Thanks again for running that test file!

JB
The parity data is distributed across all members, and the performance is a result of all disks in the set. The parity calculations and writes are what slows the throughput down from what they'd be capable of in a stripe set. But the redundancy is well worth it, and is only possible in a MP via a proper card, which the ARC-1222x is.

Honumaui

Sep 7, 2010, 05:46 PM

yeah I will run it again just to make sure :) because ya never know :) heheheh

but the read itself is good for sure

OK redoing test my machine is doing stuff already ? so a typical working days a ton of stuff running going on yada yada
88 seconds to open
saving/closing generating full composite took 30 seconds 1:23 after that for a total of 1:55

yeah just to repeat :) what you said
right now this is a raid 6 areaca 1222x seagate 750 drives they are on the OK list :)
SSSd 100 RE boot and another for scratch
I have 4 1TB hitachis in here now with raid 0 for one more layer of BU when I move data around I like extra plus I am going to put these in raid 1/0 1+0 for another tests to see if this machine becomes the layout or I take the new one ? but they are sitting their so dont count ;)

BU PM case raid 5 with 2TB drives inside and another one for Time machine

we have a server with current client files and I am sending my parents another HDD here soon that goes to another state
for if our house burns down etc.. kinda thing
the next one I might be buying some of those new 3TB seagates frys has them for $199 I have 2.87 or something of data current to send up to parents house :) yup even at 47 I use my parents house to store my off site stuff hehehehe
otherwise I would use a safety deposit box ? or trusted friends ?
I just buy discs and keep the boxes everything ? copy over pack em back up :)

my 16 bit were more in line with your times that was the 12.77 gig file though ?

Nano yeah the two cards cause I would need the two for my two backups :) that was why I said two :) I tend to speak of my stuff sometimes :)
and my RE 100 gig tap up close to that 500 so thought breathing room :) and again I need two PM ports :)

since i use two PM cases per computer :) call me paranoid but never lost data in over 30 years on computers :)

did not know the 40s were that slow ? or that fast :) depending on how you look at it so yeah the two icy dock me was thinking two SSD one boot one scratch did not make that clear my bad :)
and yeah for sure my areca is my main :)

again so many ways to configure things

JulianBoolean

Sep 7, 2010, 06:03 PM

Okay, I see two options here. One route is to go mostly all inside the box, the other is the 8 bay external solution described previously. I've sketched it out with a little little rough pricing to compare the two. From nanofrogs recent reply, it seems like two SSDs in the bottom optical bay is possible, and a third for boot won't saturate the system, which opens up the possibility for the boot in bay1, the 2x SSD for scratch in the bottom optical, then a three member raid 5 in bays 2-3-4. Then the only thing outside the box would be my two FW drives in JBOD for Back up. The other option, which would get me the smoking fast read/writes that honomaui is getting, would be obtained by way of the 8 bay external previously linked.

Option 1: Everything Inside Except for Back Up

32GB Ram
Four 8GB sticks,,, $1,500

Top Optical Bay : Leave the CD Reader in there

Empty Optical Bay : Scratch, RAID 0
2 X 40GB OWC Extreme Pro SSD ,,, $236 for the pair

Internal Bay 1 : Boot + Apps
1x 50GB OWC Extreme Pro RE SSD,,, $200

Internal Bays 2, 3, 4 : Working Data, RAID 5
3x 2TB WD RE4 (RAID Edition) ,,, $900 total for 3
Q: 4TB Capacity?

External FW : Back Up JBOD

Q: Possible saturation / throttle of internal architecture?

Est Total Cost : $2,836
+ Plus Card
+ a mounting bracket, PCIe card, Icy dock or whatever is needed
to put 2 SSD in the bottom optical drive

[..]…………

Option 2 : 8 Bay External Tower, To Get Smoking Fast Read/Writes

32GB Ram
Four 8GB sticks,,, $1,500

Internal Bay 1 : Boot + Apps
1x 50GB OWC Extreme Pro RE SSD,,, $200

Internal Bay 2 - 3 : RAID 0 Scratch
2 X 40GB OWC Extreme Pro SSD ,,, $236 for the pair

Bay 4 : Open / Expansion

8 Bay External Box : Working Files, RAID 6
Sans Digital Tower RAID ,,, $390
8x 1TB Western Digital RE Raid Edition 32 MB Cache $1,200
6TB capacity?

External FW : Back Up JBOD

Est Total Price : $3,526 + RAID Card

I think both these methods might work, as always, critque is welcome!

--Julian

JulianBoolean

Sep 7, 2010, 06:11 PM

yeah just to repeat :) what you said
right now this is a raid 6 areaca 1222x seagate 750 drives they are on the OK list :)

Cool, but still not clear on the RAID 6. Exactly how many disks on your raid 6 array? Thanks!

Honumaui

Sep 7, 2010, 06:11 PM

one thought I did not know the 40s only did 175 ? so that changes things a bit :) I have yet to get mine !!! tomorrow

Honumaui

Sep 7, 2010, 06:36 PM

Cool, but still not clear on the RAID 6. Exactly how many disks on your raid 6 array? Thanks!

8 :)

I have a lot of discs around maybe 60 or more easy ? so for me I used what I had ? I had 13 of all the same disc from a earlier raid setup ?
I would have gotten the 1TB WD RE 3 if I were starting from scratch
I would use a 120 gig for boot ? just the regular even though I have the 100 RE I got them thinking I was going to need them ? thats me :)

I am a fan of 8 discs in a external to get that extra speed ! if you are going to have a cookie HAVE A COOKIE !!! forgot where I heard that :)

nanofrog

Sep 7, 2010, 10:48 PM

Option 1: Everything Inside Except for Back Up

Est Total Cost : $2,836
+ Plus Card
+ a mounting bracket, PCIe card, Icy dock or whatever is needed
to put 2 SSD in the bottom optical drive
Add :

$624 for the RAID card (target = ARC1880i)
$129 for the MaxUpgrades adapter to use the RAID card internally
$60 for 3x Icy Dock enclosures (mount the SSD's internally)
$130 for the PM Enclosure (http://www.amazon.com/Sans-Digital-TowerRAID-TR4M-Multiplier/dp/B000YTRYEM) (4 bay unit)
$480 for 4x 2TB Green HDD's (http://www.newegg.com/Product/Product.aspx?Item=N82E16822136514&cm_re=2tb_hard_drive-_-22-136-514-_-Product) for the JBOD unit (gives you 8TB of backup capacity)
$80 for the newertech PM eSATA card (faster than the card that comes with the Sans Digital PM enclosure, though it will work)

Grand total = $4339

Performance = 250MB/s or so (based on the 1680 series)

Option 2 : 8 Bay External Tower, To Get Smoking Fast Read/Writes

External FW : Back Up JBOD

Est Total Price : $3,526 + RAID Card
Add:

$780 for the RAID card (target = ARC1880x)
$60 for 3x Icy Dock enclosures (mounts the SSD's internally)
$130 for the PM Enclosure (http://www.amazon.com/Sans-Digital-TowerRAID-TR4M-Multiplier/dp/B000YTRYEM) (4 bay unit)
$480 for 4x 2TB Green HDD's (http://www.newegg.com/Product/Product.aspx?Item=N82E16822136514&cm_re=2tb_hard_drive-_-22-136-514-_-Product) for the JBOD unit (gives you 8TB of backup capacity)
$80 for the newertech PM eSATA card (faster than the card that comes with the Sans Digital PM enclosure, though it will work)

Grand total = $4276

Performance = 535MB/s (based on the 1680 series; the 1880X should be a bit faster)

So, for $63 less, you even get faster performance. :eek: That's what more disks that cost less each can do for you. :D

JulianBoolean

Sep 8, 2010, 11:44 AM

Add :

Add:

$780 for the RAID card (target = ARC1880x)
$60 for 3x Icy Dock enclosures (mounts the SSD's internally)
$130 for the PM Enclosure (http://www.amazon.com/Sans-Digital-TowerRAID-TR4M-Multiplier/dp/B000YTRYEM) (4 bay unit)
$480 for 4x 2TB Green HDD's (http://www.newegg.com/Product/Product.aspx?Item=N82E16822136514&cm_re=2tb_hard_drive-_-22-136-514-_-Product) for the JBOD unit (gives you 8TB of backup capacity)
$80 for the newertech PM eSATA card (faster than the card that comes with the Sans Digital PM enclosure, though it will work)

Grand total = $4276

Performance = 535MB/s (based on the 1680 series; the 1880X should be a bit faster)

So, for $63 less, you even get faster performance. :eek: That's what more disks that cost less each can do for you. :D

Gentlemen, I declare this mission accomplished. It's a wrap. Have a cigar. Have a cookie!

I'm going with the plan listed above. I won't be able to afford everything all at once, but the most important thing is that I have a plan in place that I can grow into over time. I'll get the RAM first, then the SSDs and so on. I had an amazing amount of support from half a dozen members here, and every reply was really helpful, but you two really hung in there with me to the very end. I cannot thank you enough!

-Julian

Honumaui

Sep 8, 2010, 03:02 PM

no worries :) glad ya got a roadmap for your setup :)

nanofrog

Sep 8, 2010, 03:08 PM

Gentlemen, I declare this mission accomplished. It's a wrap. Have a cigar. Have a cookie!

I'm going with the plan listed above. I won't be able to afford everything all at once, but the most important thing is that I have a plan in place that I can grow into over time. I'll get the RAM first, then the SSDs and so on. I had an amazing amount of support from half a dozen members here, and every reply was really helpful, but you two really hung in there with me to the very end. I cannot thank you enough!

-Julian
:cool: NP. :)

Once you get it all put together, you'll have a system that screams, causing you to do this... :D :D :D

Khendal

Sep 8, 2010, 05:24 PM

Sorry for this stupid question...but english is not my first language...and...i don't know the meaning of scratch...could anyone of you please explain me what is it with example ?

Thanks :p

philipma1957

Sep 8, 2010, 05:27 PM

Sorry for this stupid question...but english is not my first language...and...i don't know the meaning of scratch...could anyone of you please explain me what is it with example ?

Thanks :p

kind of like a blackboard in a school it gets written on and erased over and over.

Khendal

Sep 8, 2010, 05:34 PM

kind of like a blackboard in a school it gets written on and erased over and over.

ah ok...something like inside the ram, or caches....temp file...

Thanks very much

philipma1957

Sep 8, 2010, 05:52 PM

ah ok...something like inside the ram, or caches....temp file...

Thanks very much

yes you are close. this is why some say use 2 40gb low cost ssds in raid0 as a big very fast scratch disk. you use it for great speed. ssd are good for the speed but will wear out pretty fast. thats why 2 40gb low cost ones are used. you get that speed at not too terrible a price.

JulianBoolean

Sep 9, 2010, 03:37 PM

Nanofrog -

Hey... just want to make sure I'm looking at the correct models of some of your recommendations!

Is this the correct model icy dock thingy you are recommending?
http://www.newegg.com/Product/Product.aspx?Item=N82E16817994064

Is this the "$80 for the newertech PM eSATA card (faster than the card that comes with the Sans Digital PM enclosure, though it will work) thingy?
http://eshop.macsales.com/item/Newer%20Technology/MXPCIE6GS2/?APC=XLR8YourMac09

And the Areca 1880x is not available yet, but will be soon?

Thanks again! Julian

nanofrog

Sep 9, 2010, 05:35 PM

Is this the correct model icy dock thingy you are recommending?
http://www.newegg.com/Product/Product.aspx?Item=N82E16817994064
Yes. :)

Is this the "$80 for the newertech PM eSATA card (faster than the card that comes with the Sans Digital PM enclosure, though it will work) thingy?
http://eshop.macsales.com/item/Newer%20Technology/MXPCIE6GS2/?APC=XLR8YourMac09
No, the link you posted is for the non PM enclosure model (only runs 1x disk per port = 2 disks total).

NewerTech MAXPower 6G PCIe eSATA RAID Controller card with port multiplier support (http://eshop.macsales.com/item/Newer%20Technology/MXPCIE6GRS/) = the unit that supports Port Multiplier enclosures, and the price has come down a tad to $75. :)

And the Areca 1880x is not available yet, but will be soon?
I've not called to see if pc-pitstop has it in stock, as the site doesn't say one way or the other. The cards just formally announced on August 9, 2010.

The ATTO models announced earlier, and are still hard to find, so the fact that Areca's are already available as I understand it, is a really good thing, and bodes well for the unavailable models (should be available soon). :D

I'd call pc-pitstop to see when they'll get it, so you know about when they'd be able to ship one to you (figure 1 - 2 days before they'd pack and ship it to you). But a European supplier (http://www.webconnexxion.com/raid/product_info.php?products_id=991) (scroll down to near the bottom) claims to have them in stock, so it shouldn't be too long if US suppliers are waiting/already out of their initial stock (US usually gets them first, as it's considered a larger market as I understand it).

I've checked other sources, but they're not even listed (newegg has a few models listed, but not the entire line yet; ARC-1880X = not even listed). Not sure they will carry all of them anyway, going by their past history. I presume they will only carry the models that are most likely to move the fastest (more volume per quarter = more profit).