Can you Insert Drive While Running Mac Pro ??

[fab5freddy]

Can i insert a harddrive on my Mac Pro while it is running ?

Just don't want to shock myself or the like.........

[Tallest Skil]

NO. They are NOT hot-swappable. You'd probably just fry the drive/computer.

[fab5freddy]

Oh Great! I didn't do it Yet!

Thanks for saving my computer.

[Roy]

Oh Great! I didn't do it Yet!

Thanks for saving my computer.

Leave off the "YET" part and read your manual.

"A mind is a terrible thing to waste".

[Spanky Deluxe]

Leave off the "YET" part and read your manual.

"A mind is a terrible thing to waste".

A Mac Pro is an even worse thing to waste. :p

[fab5freddy]

Sorry, i will never do that !

the word "Yet" was a mistake.

Though Apple should make them hot-swappable !

[Roy]

Sorry, i will never do that !

the word "Yet" was a mistake.

Though Apple should make them hot-swappable !


At first I thought: No Way! But the more I think about it you might be right. I have an external drive I use on my old G4 Mac Tower and I plug it in and out and never think about even shutting the G4 down.

[nanofrog]

Hot swap has more to do with the software drivers than hardware, though there is a hardware aspect to it. On windows systems, it's part of the AHCI driver. I'm not aware of this ability built-in to OS X, but is on 3rd party drivers for SATA controllers or RAID controllers.

On the hardware side, there's a chip called an Inrush Current Limiter. Basically, it's an electronic fuse to protect the PSU.

[mchalebk]

One of the problems that needs to be solved when hot swapping components is how the various pins make contact. Not all pins will make contact at the same time, so you run the risk of damaging components if things don't come together "just right". Considering that Apple is using the standard connectors that come on the hard drives, I don't think it's practical to make them hot-swappable. If you need hot-swappable drives, go Firewire or USB.

[nanofrog]

One of the problems that needs to be solved when hot swapping components is how the various pins make contact. Not all pins will make contact at the same time, so you run the risk of damaging components if things don't come together "just right". Considering that Apple is using the standard connectors that come on the hard drives, I don't think it's practical to make them hot-swappable. If you need hot-swappable drives, go Firewire or USB.
Um, No. The connectors used are backplane types, and contain both the power and data signals. Contact is at the same time for all intent, and isn't an issue. The driver support takes care of that aspect anyway (polling timings).

Hot swapping is practical, but does need a little additional attention to both the drivers and hardware (current limiting). It's rather simple actually, but does require additional development time, and components, increasing the cost.

If it's not considered necessary, hot swap capabilities will usually get skipped in favor of lower costs. FW and USB do require it, so it's present. It always seems to come down to money. ;) :p

[remmy]

Wouldn't swapping the drives mean opening the case when running which is generally bad for the computer anyway.

[nanofrog]

Wouldn't swapping the drives mean opening the case when running which is generally bad for the computer anyway.
In all likelihood, it would be open, close, remove the drive from the sled, attach the new drive, open, reinstall, and close. So No, not really, as the door wouldn't be open for any real length of time. ;)

[zmttoxics]

It is possible to hot swap sata drives but it depends on the hardware implementation (wether or not the controller / drivers allow for it - some do and some don't). In this case, the definitive answer is in the manual.

[Dzokayi]

Connect a hard drive to the lower drive bay while the machine was running. Nothing happened. It didn't mount, and disk utility didn't see it until I rebooted.

[zmttoxics]

Connect a hard drive to the lower drive bay while the machine was running. Nothing happened. It didn't mount, and disk utility didn't see it until I rebooted.

That's kind of a mute point because the G5 and Mac Pro have different SATA controllers.

[goMac]

The XServes are hot swappable...

[nanofrog]

The XServes are hot swappable...
Yes, but unfortunately, a different animal. ;)

[goMac]

Yes, but unfortunately, a different animal. ;)

But it implies it's not a software issue...

[nanofrog]

But it implies it's not a software issue...
Actually, No. :eek:

The X Serve is a rackmount system, meant for enterprise use. This implies RAID, particularly hardware implementations, which must have hot-swap capabilities. So Apple included the Inrush Current Limiters either on the board, or installed on their RAID Pro card. Ideally, they're installed on the board, just in case the RAID is in fact software based for cost savings. It also runs the SERVER version of OS X, which contains the software support.

So you get both parts needed on the X Serve. :)

[xraydoc]

That's kind of a mute point...
Moot. Moot point. Not "mute."

[Dr.Pants]

In response to the physical implications of hot-swapping, SATA has the grounds make contact before the pins... FYI. I can see how if everything did not line up correctly in IDE that would screw things up.

As for the Inrush Current Limiter... Isn't that just a circuit with a resistor and a big inductor? Physics happens to be grinding my gears with finals coming up, and would really like to rattle off a part number to irk my physics prof. :p

[Darth.Titan]

NO. They are NOT hot-swappable. You'd probably just fry the drive/computer.

LOL
Not likely. Sticking the drive in with the computer running will not cause any problems (yes I have done it) however the drive would not mount until you reboot because (as has already been stated) the internal drive bays are not hot-swappable.

[nanofrog]

In response to the physical implications of hot-swapping, SATA has the grounds make contact before the pins...
Yup. :)

I was primarily thinking in terms of not being able to insert on an angle as is possible with IDE. Not nearly the time delay between contacts.

FYI. I can see how if everything did not line up correctly in IDE that would screw things up.
This is definitely an issue with IDE.

As for the Inrush Current Limiter... Isn't that just a circuit with a resistor and a big inductor? Physics happens to be grinding my gears with finals coming up, and would really like to rattle off a part
number to irk my physics prof. :p
There's a few ways. The simplest and most common, is either a fixed resistor or thermistor.

But there are IC's used as well (PFC). Here's an example (NCP 1650DR2/G (http://www.onsemi.com/pub_link/Collateral/NCP1650-D.PDF)) by On Semiconductor. :D

Just print out the data sheet for your professor. ;) :p

[Dr.Pants]

The simplest and most common, is either a fixed resistor or thermistor.

Epic facedesk moment. Current is inversely proportional to resistance (V=IR) - for some reason, I was thinking that an inductor would be a better on the whole as the current is then based off of an exponential function. However, it approaches a limit as time goes to positive infinity, so the actual voltage across the inductor grows fairly rapidly and then slows. lim time->negative infinity was what I was thinking...

[/most random physics rant ever]

Oddly enough, I'm only half-sure that my physics professor would understand that datasheet. Electrician in the Navy, had to learn circuits the right way (we learn positive charge flows; protons, anyone?) and did her postdoc with string theory. Electronics may not be her thing, but I would like to get the idea of what each pin does (and the purpose therof) before I actually wing out the stapler with a group of pages which many answers may reference with the words "see attached" :D

Sorry to move off-topic, its late and I've been up too long writing a lab.

[nanofrog]

Epic facedesk moment. Current is inversely proportional to resistance (V=IR) - for some reason, I was thinking that an inductor would be a better on the whole as the current is then based off of an exponential function. However, it approaches a limit as time goes to positive infinity, so the actual voltage across the inductor grows fairly rapidly and then slows. lim time->negative infinity was what I was thinking...

[/most random physics rant ever]
I've had my moments, so you're not alone. :D

Oddly enough, I'm only half-sure that my physics professor would understand that datasheet. Electrician in the Navy, had to learn circuits the right way (we learn positive charge flows; protons, anyone?) and did her postdoc with string theory. Electronics may not be her thing, but I would like to get the idea of what each pin does (and the purpose therof) before I actually wing out the stapler with a group of pages which many answers may reference with the words "see attached" :D

Sorry to move off-topic, its late and I've been up too long writing a lab.
The schematics should help immensely. ;)

As you can see, there's a lot more to it than just a simple current limiter like a thermistor. ;)

[eyecool]

Man, I've been running my MP with the panel open, plugging and pulling my wife's fried iMac drive in to hard drive bays 3 and 4.

I've also got a USB connector for hard drives plugged in. Turning it off and on and off and on trying to figure out how sizzled this iMac drive is. It was BURNING hot when I pulled it out. Second time this has happened. First time it was just out of warranty, so I had no fear of taking it apart.

Since I've been searching for help with this drive, I was already logged in to the forums. One suggestion was putting the drive in the freezer, then connecting it. The longer it takes the drive to heat up, the more data u can pull off. It's the only thing I've found (so far) that works.

So here I was, about to plug a freezing cold, seagate 500 gb, 7200.10 in to bay 3 of my dear ol' eight core big mac when I figured I'd search to make sure drives were hot swapable.

THANK GOD I READ THIS THREAD! MY MAC IS NOT DEAD YET! Although I did have to reboot 6 times today, and boy, do I hate rebooting. This beast is always on.

Time to put the cover back on, get it off the CARPET and call it a night. Mac Pro hard drive are definitely not plug n play "hot swapable".

[lixuelai]

Lol you cannot fry it. SATA is hot swappable. However the software has to support it. Only bad thing would be if you pull a running drive and then it is your data not the hardware getting hosed.

[Tesselator]

Yeah, I think nothing would be damaged either but why chance it when there is no purpose. Nothing is gained by plugging one in while the system is running - you still need to restart. Eject works fine instead of unplugging so... nothing gained - no reason to try it.

[nanofrog]

Lol you cannot fry it. SATA is hot swappable. However the software has to support it. Only bad thing would be if you pull a running drive and then it is your data not the hardware getting hosed.
Hot Swappable and Hot Pluggable aren't the same though. In the case of USB for example, it is. But not with SATA.

If the power portion is capable, the connector is a little different. Pins 4 and 12 are longer (grounds), and there's a circuit installed to limit the inrush current. Usually on the board, as most PSU's don't support it by themselves. That's not the case for the '06 - '08 models, and appartently, the '09's either. :( The XServe is the exception from Apple I'm aware of.

So it has Hot swap capabilties built in for the data (cabling), but doesn't have the software support in the client edition of OS X. Effectively useless.

And the power section doesn't have the circuits needed for Hot Pluggable devices either. So there is the possibility to damage the PSU. :eek: Now the size of the PSU used helps, and given the average power usages posted, I don't think it should be worried about, but specific use can change this (really high avg. draw).