Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
![]() |
World Community Grid Forums
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
No member browsing this thread |
Thread Status: Active Total posts in this thread: 63
|
![]() |
Author |
|
Curly2
Cruncher Joined: Nov 24, 2008 Post Count: 10 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Result DSFL_ 00000050_ 0000004_ 0365_ 1 uses more than 1GB memory.
----------------------------------------This is really surprising as the hardware requirements are 250 MB memory available for DSFL tasks. Is it intended to have such a high memory usage? Edit: OS and Version answered in this post . ![]() [Edit 1 times, last edit by Curly2 at Oct 31, 2011 1:13:55 PM] |
||
|
genes
Advanced Cruncher USA Joined: Jan 28, 2006 Post Count: 132 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Hmm, over the last day or so, I've found all three of my computers unresponsive, with maybe about 50% of CPU time unaccounted for (on quads, only two of the four tasks appearing to be getting CPU time). Didn't think to look at the memory usage (but I will next time). After shutting down BOINC manager and restarting it, everything was back to normal.
Came here to see if others were having issues, don't know if this is it, but it might be related. |
||
|
genes
Advanced Cruncher USA Joined: Jan 28, 2006 Post Count: 132 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Okay, it is the same issue. On this machine, a quad with 4GB of memory, I have four tasks running, three of which are using over 900MB, the other using over 600MB. I had to shut BOINC down in order to type up this post.
Looks like a bug here, or we will have to have a limit on DSFL tasks like CEP2. |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Fast conclusions ;>) Could be the specific batch/target as presently mine used 36MB RAM *peak* after 13:30 hours and 87% progress at 97% efficiency. The VM is 53 M AND my usable RAM is limited to 2.2GB when in use, which is the control you can use to let BOINC pause tasks when u need it. There's also the CPU load control % I use under Linux as else jobs go bad [time out] during use when the system is very busy *for me*.
Question: Running at 100% CPU time? What OS? What Client version? --//-- |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The same here, unable to work with my PC regardless of restarting boinc Client,
issue comes back after short period of time. Disabled the Project for now. |
||
|
widdershins
Veteran Cruncher Scotland Joined: Apr 30, 2007 Post Count: 674 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Same here, I have a quad that apparently has 8 processes running 1 taking just over 1Gb and the others just under. Now this is quite a feat as it's a q6600 and doesn't have hyperthreading. It seems to be starting some processes, running a bit then switching to another.
----------------------------------------Edit: Just checked again and now there's 10 processes, 4 taking up just over 1/2 Gb each, the rest around 1Gb. ![]() ![]() Looks like Mr Quaddie is going on a WU diet, only 1 unit at a time. I'd like my HDD to survive more than a week. Hopefully it's just a difficult batch of WU's and normal service will be resumed soon. Edit2: OS is Kbuntu 8.04LTS 64bit, Boinc is 5.10.45, and yes it was running Boinc @100% on all cores until now. [Edit 2 times, last edit by widdershins at Oct 31, 2011 12:13:36 PM] |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
BOINC is a little [major] bit stupid, but what else can it do if it only got DSFL in stock for WCG. It will try to fit jobs in available memory so tries a few then when it finds it's not fitting tries another combo. Set the memory use lower in the prefs and it will try to run 3 and probably much faster.
Whilst only one poster has shared it's Target 50, batch 4 that has this problem. Client Version? OS? HDD will not suffer that much if defragged once in a while and a fixed, contiguous spaced VM (default in Linux). --//-- |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Oh Yes, is LAIM ON? If not the job alternations will cause them to unload constantly, loosing progress.
--//-- |
||
|
GreatWelshOptimist
Cruncher Joined: Jan 20, 2009 Post Count: 3 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Have the scientist behind this project just released a second workunit type?
The reason I'm asking is that when I look at the processes I currently have running, I seem to have 2 different DSFL workunit types. There seems to be the standard wcg_dsfl_6.19_indows_intelx86 workunit, as well as a wcg_dsfl_vina_6.19_windows_interx86 workunit which I can't remember having seen before. And it is the latter workunit that is hogging the system memory, currently 800MB on the machine that I am typing this on. Mine is target 50 batch 08. |
||
|
krakatuk
Advanced Cruncher Germany Joined: Oct 3, 2008 Post Count: 141 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Same here, I have a quad that apparently has 8 processes running 1 taking just over 1Gb and the others just under. Now this is quite a feat as it's a q6600 and doesn't have hyperthreading. It seems to be starting some processes, running a bit then switching to another. Strange, but I'm seeing similar weird BOINC behaviour on my i7 2600K with CEP2. On Saturday I was playing around with cache settings and accidentially have configured BOINC to load 5 days of cache instead of 0.5... I'm running CEP2 only on this box, so I got 240 CEP2 WUs directly after that. But it's not the issue... The issue is, that BOINC cannot decide itself which tasks tu run. It starts 8 WUs, runs for an hour and then gives higher priority to another 8 WUs and jumps on them. I've seen up to 24 WU's being left half finished same time. You can imagine how much memory 24 CEP2 WU's need ![]() Still the box seems to get on with that, so I left them running as BOINC wishes, but the behaviour is more than strange. My system has Windows7 x64, 8Gb RAM. All WUs have Nov 8th as deadline, so I don't see the reason for BOINC to feel stressed. Edit: sorry for OT. It has definitely nothing to do with DSFL and original topic... ![]() [Edit 1 times, last edit by krakatuk at Oct 31, 2011 1:08:19 PM] |
||
|
|
![]() |