Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Help the MacRumors SETI Team get on the Top 200 Lists!
- Thread starter simX
- Start date
- Sort by reaction score
I'm not familiar with SETI anymore (was, but that was a while ago) but the client version of FAH allows you to set the frequency at which the program will "checkpoint" itself and save it's work to disk. I think the default is 15 minutes. The checkpoint is the place where the program will restart folding if for some reason it is terminated. So you can control to some extent the disk activity. Incidently, powering down disk drives has historically been done for power saving reasons, not MTBF. I don't think that has changed, but could be wrong.
As for how much CPU time you need for a WU, it depends on what kind of computers you have. Faster is obviously better. What do you have available?
The netboot question has come up before, I don't recall what the resolution was, if there was one. From the Stanford FAQ
Can I run both the screen saver and console version at the same time? What happens if I run two console versions at the same time?
ONLY if you install in different directories. Also, DO NOT DO NOT DO NOT just copy the files from one directory to each other. This will cause our server to get confused, and 1) you won't get credit for the work you do, and 2) it will be no use for science. Instead, run the install program twice into each directory. If you have already copied the program into multiple directories and are trying to run it, find the client.cfg file and delete the "userid=xxxxxxxx" line. Next time it connects to F@H, it should reset the client.cfg file, and all should be well.
While this is talking about multiple installations on a single machine, the implication is that the servers need something in the client.cfg file to tell individual computers apart. Thus, if every one of your machines starts with an identical boot image AND you can't have unique FAH directories with unique client.cfg files, you may end up "confusing their server".
Since my G4 never crashes (I explained to my Mac that crashes are against my policy, and it politely complied), I can presumably set it to checkpoint every 24 hours or more and not worry about disk usage.
It stands to reason that starting up and spinning down is part of the wear and tear on a disk drive, and is actually worse than leaving it running for short periods in between times of usage. I imagine it is like a lightbulb. You get a certain number of hours out of it. If you leave it on for hours, then off for hours, you probably get the most life out of it. If you leave it on all the time, you'll get less. And if you turn it on and off constantly, you'll get less. Since I don't have a notebook Mac, I'm not worried about power consumption by the disk drive, which is another tradeoff for iBook and PowerBook users.
Dual 1.25GHz Power Mac G4. I can do a SETI unit in 3 hours.
Thanks. I run two CLI SETIs and understand its processing and ID and status file usage. Folding is probably similar in this respect. I think the approach with the best chance is if I try to have the data file on each student computer's local disk, not the netboot image and not the home folder on the server.
I don't know. I've been asking more questions than I can answer and my assumptions are only guesswork. Maybe macintouch has some info about this topic.
The disk drive issue is probably a good subject for a new thread in the hardware forum. After searching for previous threads about the topic, of course.
The disk drive issue is probably a good subject for a new thread in the hardware forum. After searching for previous threads about the topic, of course.
re:netboot -- this is a linux example. (i never used netboot, it's probably wildly structurally divergent from nfs-mounted home dirs. anyway)
i'm running 34 fah clients from one home dir (i have a .fah folder and 34 fah# folders inside it).
my home dir is mounted by the machines automatically when needed (nfs). so basically, my home dir is mounted on 17 linux boxes full time.
i thought of running on local drives, but it's so much easier to keep track of the linux boxen this way (since they don't run web servers, i can't do the 'ln -s' into the web dir trick i use on the macs to make the fah data more available remotely. i can run shell scripts, aliases and cron from any one linux box (like mine) and process data for all the procs. (i still had to rsh and set up 17 crons to do the actual fah start scheduling on each machine)
each proc has a unique id because i set them up individually (i make the dir, copy the client binary only and then run ./FAH4Console-Linux.exe -local -configonly and set the crontab for each proc)
RE:HD Access -- i think uptime and sleep/no sleep is a personal preference. i've had machines running for years with no failures - but sometimes not (i do back up my home dir nightly now).
my dual 500 G4 has run 'continuously' for 3 1/2 years. it's been running fah 24/7 since early december at full bore. the G5, too. and my work mac. in fact, work macs are not allowed to sleep. well, either are home macs. and i run them all hard, all the time. i'm sick.
here's the g4's uptime - i added some software updates recently, but you can see each proc is fully utilized and it has been up for 12 days.
bat:~ admin$ uptime
18:32 up 12 days, 4:54, 2 users, load averages: 2.00 2.00 2.00
before you accept my word, remember that i'm not afraid to lose _some_ data, i do back up alot of stuff and frequently, and i do all my own hardware swapping/repairing. historically, i've constantly had to reinstall my OS on my main mac because i get too jiggy with the cutting edge stuff. (yet the replacement G5 apple gave me has somehow survived all my schemings so far - knock on wood).
- hope this is of some good to you.
i'm running 34 fah clients from one home dir (i have a .fah folder and 34 fah# folders inside it).
my home dir is mounted by the machines automatically when needed (nfs). so basically, my home dir is mounted on 17 linux boxes full time.
i thought of running on local drives, but it's so much easier to keep track of the linux boxen this way (since they don't run web servers, i can't do the 'ln -s' into the web dir trick i use on the macs to make the fah data more available remotely. i can run shell scripts, aliases and cron from any one linux box (like mine) and process data for all the procs. (i still had to rsh and set up 17 crons to do the actual fah start scheduling on each machine)
each proc has a unique id because i set them up individually (i make the dir, copy the client binary only and then run ./FAH4Console-Linux.exe -local -configonly and set the crontab for each proc)
RE:HD Access -- i think uptime and sleep/no sleep is a personal preference. i've had machines running for years with no failures - but sometimes not (i do back up my home dir nightly now).
my dual 500 G4 has run 'continuously' for 3 1/2 years. it's been running fah 24/7 since early december at full bore. the G5, too. and my work mac. in fact, work macs are not allowed to sleep. well, either are home macs. and i run them all hard, all the time. i'm sick.
here's the g4's uptime - i added some software updates recently, but you can see each proc is fully utilized and it has been up for 12 days.
bat:~ admin$ uptime
18:32 up 12 days, 4:54, 2 users, load averages: 2.00 2.00 2.00
before you accept my word, remember that i'm not afraid to lose _some_ data, i do back up alot of stuff and frequently, and i do all my own hardware swapping/repairing. historically, i've constantly had to reinstall my OS on my main mac because i get too jiggy with the cutting edge stuff. (yet the replacement G5 apple gave me has somehow survived all my schemings so far - knock on wood).
- hope this is of some good to you.
I wouldn't worry about the mean time between failure. If you check the websites of the big disk makers for mean time between failures (MTBF), you'll see numbers like 1.2 million hours. On average, these things break every... lessee, there are 8760 hours in a year, so you can expect your average enterprise disk to head for the great beyond once every 137 years of continuous operation. Of course, if you have 137 drives, you can likewise expect one of them to fail each year, which is one of the reasons why reliability is such a big deal to the disk drive makers.
Sure, these are numbers for enterprise drives, but for your personal computer running 24/7 for 5 years, you'll only put about 44000 hours on it. Chances are though, that ancient dog-o-a-puter will be gone well before then! Certainly you won't be able to get by with a mere 250 GB in a year or two anyway...
Just make dadgum sure you don't do anything that blocks up or otherwise impedes the cooling. That has a tendency to be fatal to electronics of all sorts, including disk drives.
Damn -- I was biased?
Drat I was trying to just inform... not flame...
Sorry SETI people!
wdlove -- You can turn off your machine... or even let it sleep... I mean I have my machine running for... 8 days currently but obviously there are problems with not restarting or letting your machine cool off.
Anything above 9 days for me and I let my machine atleast cool off for an hour.
Its a good machine to me.
--MrMacMan
Originally posted by MrMacman
Damn -- I was biased?
wdlove -- You can turn off your machine... or even let it sleep... I mean I have my machine running for... 8 days currently but obviously there are problems with not restarting or letting your machine cool off.
Anything above 9 days for me and I let my machine atleast cool off for an hour.
Its a good machine to me.
--MrMacMan
Then in your opinion, I should do a shutdown of my Mac once a week? That would be easier to remember. I just want to do what is best for my Mac and Folding.
Off topic MrMacman, I have noticed recently that you are capitalizing all 3 M's. Which do you prefer?
I have 170+ UW on the G4 1.25 PB, and another 260+ WU on the Bondi iMac and 7600 PowerMac(120mhz baby!).
For the fun of it-
G4 1.25ghz = 9 hr WU
G3 233mhz = 26 hr WU
604e 120mhz = 40+ hr WU
I was about to transfer over my UW from my previous dead team that I joined way back when. But I looked up my team for the fun of it...My Bungie Halo teal, started 4 years ago, is now ranked #137 in the Clubs
sorry guys..maybe I'll come in at the last minute to take the gold with you
Tyler
Earendil
For the fun of it-
G4 1.25ghz = 9 hr WU
G3 233mhz = 26 hr WU
604e 120mhz = 40+ hr WU
I was about to transfer over my UW from my previous dead team that I joined way back when. But I looked up my team for the fun of it...My Bungie Halo teal, started 4 years ago, is now ranked #137 in the Clubs
sorry guys..maybe I'll come in at the last minute to take the gold with you
Tyler
Earendil
wdlove -- Yeah I mean feel free to restart your machine or shut it down.
Your overall computer and OS needs a restart because I don't know, they become sluggish.
OffTopic -- I don't know really... It was once 'Mrmacman' just for lazyness, but by the time I registered it was 'MrMacman' but finally just for the cool look it has become 'MrMacMan'... Chose any, heck you could even call me Jeremiah, Miah or even Jeremy, I don't care.
(MrMacMan is fine)
Earendil -- Too bad, the SETI team would love your support!
Please reconsider!
Your overall computer and OS needs a restart because I don't know, they become sluggish.
OffTopic -- I don't know really... It was once 'Mrmacman' just for lazyness, but by the time I registered it was 'MrMacman' but finally just for the cool look it has become 'MrMacMan'... Chose any, heck you could even call me Jeremiah, Miah or even Jeremy, I don't care.
(MrMacMan is fine)Earendil -- Too bad, the SETI team would love your support!
Please reconsider!
Originally posted by Stelliform
The gap has widened to 58000, I had to shut down the computers I was borrowing
I have to keep in mind how different the rules are regarding teams in SETI versus Folding. (In folding if you switch teams your points stay with the team, in SETI the points stay with you.) This accounts for the wild fluctuations in the SETI points.
With that being said I see the gap has narrowed to 51500. Macnn isn't responding at the moment, but I know we added a new person a day or so ago. 200 here we come!
Currently our gap to break top 200 stands at only 45000!
Also we just added a new member. Welcome Brushy! Thanks for the additional 1400 points!
Also we just added a new member. Welcome Brushy! Thanks for the additional 1400 points!
Holy Work Units Batman! Engseld has just joined the team and added an amazing 4500 points!!!!
I would give a new state of the reaching 200, but unfortunately Seti top 200 stats are down now.... Unless anybody knows where else I can get that list....
Look at that graph jump!!! Thanks Engseld!!!!!
I would give a new state of the reaching 200, but unfortunately Seti top 200 stats are down now.... Unless anybody knows where else I can get that list....
Look at that graph jump!!! Thanks Engseld!!!!!
Attachments
We currently are only 40700 WU away from the top 200! And we have just added a new member who is new to SETI also. Welcome Think Mac!
By the way, is anybody out there reading or caring? I am feeling lonely in this SETI thread.
I know Rower is plugging away with his two machines. He has a rate of 2.7 WU a day. Thanks for the help!
By the way, is anybody out there reading or caring?
I am feeling lonely in this SETI thread. I know Rower is plugging away with his two machines. He has a rate of 2.7 WU a day. Thanks for the help!
Stelliform said:We currently are only 40700 WU away from the top 200! And we have just added a new member who is new to SETI also. Welcome Think Mac!
By the way, is anybody out there reading or caring?
I know Rower is plugging away with his two machines. He has a rate of 2.7 WU a day. Thanks for the help!
It's not much, but I'm glad to help out.
Stelliform said:
For the top 200 you can look here:
http://statsman.ww-testsites.co.uk/setistats/html/
And obvious MacNN for the rest:
http://teamstats.macnn.com/fold/stats.php?TID=3446&page=b5
Awww, you guys do care..... <sniff>
Rower and Doctor Q, out of 110 people actively producing for SETI MR, you guys are ranked 32 and 35 based on production. So evey little bit makes a big difference for the SETI team.
MrMacMan, thanks for the link, that is a new one for me, but unfortunately they do the top 200 overall teams. We are struggling to make the top 200 clubs list.
In other news, we have added yet another member!! Welcome Mark Borland! He brings another 379 WU to the team. That means in the last month new members have brought in 7500 WUs just by joining our team! Anybody out there who might have even given up on SETI and they have WU lingering, please join Team MR. We will put them to good use...
Rower and Doctor Q, out of 110 people actively producing for SETI MR, you guys are ranked 32 and 35 based on production. So evey little bit makes a big difference for the SETI team.
MrMacMan, thanks for the link, that is a new one for me, but unfortunately they do the top 200 overall teams. We are struggling to make the top 200 clubs list.
In other news, we have added yet another member!! Welcome Mark Borland! He brings another 379 WU to the team. That means in the last month new members have brought in 7500 WUs just by joining our team! Anybody out there who might have even given up on SETI and they have WU lingering, please join Team MR. We will put them to good use...
For SETI, I think you mean http://teamstats.macnn.com/seti1/stats.php?TID=3446&page=b5MrMacman said:
Looks like we're gaining on the Top 200 spot... right now, we're in need of just a smidge over 40000 work units to put us up on the Top 200 Clubs list (40036 WUs needed, to be exact).
You guys really should put info about the MacRumors SETI team in your signatures so that people know about it! Publicity can only help us.
You guys really should put info about the MacRumors SETI team in your signatures so that people know about it! Publicity can only help us.
I think I have a bit of bad news. It appears that 200 won't be as easy as we hoped. It looks like a couple of groups have bumped ahead of us. We are now 46076 WU away from the 200th spot. I think we need to beef up production among our members to make the 200. (not to mention get some more members!)
So if anyone of you knows anyone running SETI (not for MR), tell them to ditch their old team, and join us!
Also I am posting this so we can use it as a benchmark. According to MacNN, we have produced 345 WU's in the last 24 hours. Our average rate is 348.84.
Lets see how much we can bring that up in the next week! I am also thinking about having a Super WU Weekend. I am going to see how many computers I can get to run at least one WU for me this weekend. (Friends, stuff like that...)
Keep up the WU production!
I think we need to beef up production among our members to make the 200. (not to mention get some more members!) So if anyone of you knows anyone running SETI (not for MR), tell them to ditch their old team, and join us!
Also I am posting this so we can use it as a benchmark. According to MacNN, we have produced 345 WU's in the last 24 hours. Our average rate is 348.84.
Lets see how much we can bring that up in the next week!
I am also thinking about having a Super WU Weekend. I am going to see how many computers I can get to run at least one WU for me this weekend. (Friends, stuff like that...)Keep up the WU production!
"Imagine waking up one day to find your Mac has solved a vexing scientific problem. While the cure to cancer, super-efficient solar power and ending world hunger are a ways off, you can combine your computing resources using Xgrid and help usher in a new era of biological breakthroughs, rocket science and advanced models of scientific phenomena"Stelliform said:
Apple > Xgrig
Seems that I heard that Xgrid is workable with SETI. Does anyone know if this is true?
Doctor Q said:
I could have sworn I posted the SETI one... not the folding one...
Huuhh... weird.
I haven't heard that the Xgrid works with SETI....