| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: The Clean Energy Project - Phase 2 Forum
Thread: Hello! Only 1 task at a time? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 10
|
|
| Author |
|
|
trashcan7
Cruncher Joined: Jan 17, 2011 Post Count: 5 Status: Offline |
Hello! I'm new to all this so forgive me if I seem like a noob :D
It seems that the Clean Energy Project only gives one task at a time- is it possible to change this? I'm also in another project, the Clean Water one, which seems to send 15 tasks at a time. And as a result, the single clean energy project is buried down in terms of priority. Can I change this as well? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi,
----------------------------------------You'll need to give the secret password with correctly mimicked sign... noobs don't likely know them ;>). After, go to the top of the CEP2 forum where there is a sticky post with the meaningful title [not] "New Project Setting", which explains the way you can select how many you'd like to have on the machine at any one time 1-16 or No Limits. Warning: This research app is HEAVY which is why the default is to only have one on a client! Slowly increase the number until you reach the right mix so your computer wont fall over in inefficiency throwing around gigabytes of data to the disk constantly swapping it out of memory. ttyl [Edit 1 times, last edit by Former Member at Jan 21, 2011 8:16:40 AM] |
||
|
|
Jim1348
Veteran Cruncher USA Joined: Jul 13, 2009 Post Count: 1066 Status: Offline Project Badges: 
|
An SSD helps a lot when running more than one CEP2 task at a time, and a Ramdisk helps even more if you have enough memory. But I seem to be in the situation where running 4 CEP2 tasks at a time on my quad-core often (but not always) causes a crash, even though I am now using the DATARam Ramdisk. Limiting the number of CEP2 tasks to 3, so that at most three are running (and usually only one or two at a time) seems to solve this problem for me. I have no idea why that is, though it almost seems to be a problem internal to the quad-core itself. I expect the motherboard stability is also questionable under such a high load, and it was probably never tested very much in such an environment.
----------------------------------------[Edit 1 times, last edit by Jim1348 at Jan 21, 2011 4:53:16 PM] |
||
|
|
trashcan7
Cruncher Joined: Jan 17, 2011 Post Count: 5 Status: Offline |
Thank you for the responses!
I, too, have a quad-core, but I think I'll limit it to 1 or 2 CEP2 tasks at a time. Do you have an answer to my second question? I'm also in another project, the Clean Water one, which seems to send 15 tasks at a time. And as a result, the single clean energy project is buried down in terms of priority. Can I change this as well? So yeah, my computer is usually doing 4 Clean Water tasks at a time, and has to get through all the other Clean Water tasks before the CEP2 task on the queue. Any suggestions? |
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 3010 Status: Offline Project Badges: 
|
Hi trashcan7,
----------------------------------------You can try 'forcing' a CEP2 WU to start - that is, by suspending ALL the tasks which are in a "Ready to Start" (and all WU's - if any exist, in a "Waiting to run") state, "Resuming" the 1 CEP2 task you want to run and then "Suspending" one of the tasks which is currently "Running". This should have the effect of starting the CEP2 task in question. After you've then "Resumed" all the tasks which you suspended, it'll leave you with a task that's "Waiting to run" - which, will continue after the next WU has completed.  |
||
|
|
Dark Angel
Veteran Cruncher Australia Joined: Nov 11, 2005 Post Count: 728 Status: Offline Project Badges: 
|
The water units aren't running at a higher priority. Unless there's a comment next to the unit listing in the BOINC manager identifying it as "high priority" (which is unusual and not something you can change) then it's first in, first crunched.
----------------------------------------As for systems crashing when allowed to run four at a time ... my GUESS is that it has something to do with the very high swap file/partition usage of these units. I don't know why but they page a LOT of data in and out of memory. Me, personally, I wouldn't install BOINC on a SSD if I was going to run CEP2. I'd put it on a regular hard drive and move my swap file/partition there as well to save chewing through the limited number of disk writes the SSD can handle.  Currently being moderated under false pretences |
||
|
|
anhhai
Veteran Cruncher Joined: Mar 22, 2005 Post Count: 839 Status: Offline Project Badges: 
|
trashcan7,
----------------------------------------Here is some advice from Sekerob, Increase your cached number by the factor the jobs run shorter and you'll likely maintain the same "concurrency". Could probably plot that in a curve using 1 core and cache size and then multiply by the number wanted to run concurrent. It seems in practice that the scheduler is tuned to priority backfill any CEP2 shortfall on the next work-fetch calls. Sample: Cache 1.5 days. Run time 8 hours: 36 hours / 8 = 4.5 * 2 Cores (half of quad), makes it 9 **. Has to be mixed with sciences that are not limited in supply. Not tested this, but over time this could balance out. My 1.2 day cache worked with a setting of 5, to get a quasi 2 out of 4 on the quad, but that was with longer running work units. Then had to cut it to 4, but that led also to periods that no CEP2 ran, so then just manually kicked them from the rear (but only when checking in a few times a day), which then on completion hastened the backfill again. In his example, it was 8hrs to complete a CEP2 WU and 1.5 days cache (36 hrs). You will have to adjust it to your own machine.  |
||
|
|
trashcan7
Cruncher Joined: Jan 17, 2011 Post Count: 5 Status: Offline |
The water units aren't running at a higher priority. Unless there's a comment next to the unit listing in the BOINC manager identifying it as "high priority" (which is unusual and not something you can change) then it's first in, first crunched. Sorry for the miswording. I meant to say that the water ones are just first in the queue, since they download 15 at a time while CEP2 only downloads 1 at a time. Thanks for the advice everyone, I was hoping there'd be a simple fix, but I guess not. @gb009761 - Thanks, this is what I'm doing for now, I guess. I was hoping for something that would require minimal maintenance of the tasks, but this works. @anhhai - O.O That went mostly over my head, but I think I get the gist of what he's saying. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
tashcan7,
To clarify the 15, that is the maximum sent to a client after a work call for X seconds of tasks. If the call was for instance to get 230,000 seconds and the job is 15,000 seconds, you'd get the maximum. Once the cache is filled, you'll get only a trickle, so you'll get things like: Request 1,500 seconds, send 1 task. This is when the CEP2 start mixing in. Soon as one is completed and reported the client will fetch a new one, if there's room per your cache (Additional buffer) setting. 1.00 day is equivalent to 86,400 seconds of cached work, per core. A quad would thus buffer on 1 day 4 times 86,400 seconds. Usually a little bit more since tasks never are sized exactly by X seconds. The client will ask for work to fill ''at least''' that amount of work as per the cache setting, which includes the time left on the jobs actually running. --//-- |
||
|
|
Jim1348
Veteran Cruncher USA Joined: Jul 13, 2009 Post Count: 1066 Status: Offline Project Badges:
|
As for systems crashing when allowed to run four at a time ... my GUESS is that it has something to do with the very high swap file/partition usage of these units. I don't know why but they page a LOT of data in and out of memory. Me, personally, I wouldn't install BOINC on a SSD if I was going to run CEP2. I'd put it on a regular hard drive and move my swap file/partition there as well to save chewing through the limited number of disk writes the SSD can handle. I am now using a Ramdisk, which has plenty of write speed and essentially unlimited endurance. By placing the BOINC data folder on the Ramdisk nothing touches the SSD from any of the WCG projects. But I agree that it would be best to keep CEP2 off the SSD, at least if you are trying to run 4 CEP2 tasks at once. I don't think running one will be much of a problem. And I think that the newer SSDs have pretty good lifetimes (check out the Intel G3s when they get here) and I expect with a quad-core they would last for several years, probably as long as a spinning drive.[Edit 1 times, last edit by Jim1348 at Jan 24, 2011 3:51:16 PM] |
||
|
|
|