| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: 10 cores running 24/7 = < 8 cpu days / day ??? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 23
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
HELP - I am running 10 cores 24 hours per day but not getting 10 days of CPU time per day - can anybody help me - PLEASE !!!
months ago I was getting a much higher turn around! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
That's rather sparse information with which to assist you! How about listing the initial event log entries following a reboot?
Have you had a recent upsurge in PVal/PVer temporary results that would reduce the apparent CPU time until they turn valid? How many machines are involved? Which WCG projects? If CEP2 only, have you restricted the number of workunits per host in the device profile? Are all cores busy on WCG at any one time in Task Manager? In general, read and act on the WCG FAQ BOINC: How to enforce Maximum Output for a device and the WCG Help item How do I get the software to use 100% of the available CPU? (in the WCG Help Troubleshooting section). |
||
|
|
OldChap
Veteran Cruncher UK Joined: Jun 5, 2009 Post Count: 978 Status: Offline Project Badges: 
|
My experience has been that the mix of wu's, the OS and the combined efficiency of these on your hardware all chip away at the target of 1 day per day per core.
----------------------------------------That efficiency can be observed real time if you use boinctasks Using Linux for example, I see that FAHV insists on messing with any settings that you may have for scheduling which in turn means that a certain percentage of your cpu time is then used by this re-scheduling. Where you are running alongside a wingman your results will not necessarily report when you return them so you can expect large fluctuations in time on a daily basis but over time this should even out. Take a look at your "device history and statistics", put those numbers through a spreadsheet and you will find the average for your rig. It is easy to look at that page and, like me, see low numbers and start worrying, but the spreadsheet tells it how it really is. To put things in context my big rig running mostly FAHV/MCM1 gets 30.77 days a day out of a possible 32 averaged over 30 days  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
....and you do not say which projects you are running?
 |
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 3010 Status: Offline Project Badges: 
|
Also, what else (if anything) have you got running on your rigs?, as all unnecessary tasks can be taking small, but significant CPU cycles away from crunching.
----------------------------------------Another factor to take into consideration, is any heat issues that may be arising - have you done a 'dust bunny' clean recently?  |
||
|
|
branjo
Master Cruncher Slovakia Joined: Jun 29, 2012 Post Count: 1892 Status: Offline Project Badges:
|
Hi Mark,
----------------------------------------Check your returned results' CPU times vs. Elapsed times. Calculate the percentage of several of them and if the averages are above 95%, it is OK. If they are significantly lower, you have to check your system and find out what program is using high share of CPU and disable it. My win rig, 100% dedicated to BOINC if efficient around 99-100%. My Mac, which I use as a computer 4 - 8 hours a day, has (average daily) efficiency around 97%. So, if your efficiency is really only 80%, it is too low and you have to find the reason. Cheers   Crunching@Home since January 13 2000. Shrubbing@Home since January 5 2006  [Edit 2 times, last edit by branjo at May 31, 2014 1:05:39 PM] |
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges:
|
HELP - I am running 10 cores 24 hours per day but not getting 10 days of CPU time per day - can anybody help me - PLEASE !!! Mark, with the way you are stating it the lazy/easy answer is "this is normal!". Even if nothing else than the operating system (OS) is running in your machines WCG work will get only 24 hours less the CPU time used by the OS. months ago I was getting a much higher turn around!  If the difference is more than very few percents we need more detailed information to have a chance to give you useful answers. First, give us the total CPU time you get on a long enough period of time, say at least 14 consecutive days. Next tell us which projects you are running: some applications (e.g. CEP2) are well known as systematically losing a significant amount of CPU time at each checkpoint. Read you later, hopefully. Jean. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I'm only running 1 project - MCM:
06/01/2014 0:000:04:02:41 803 1 05/31/2014 0:005:08:15:37 24,001 27 05/30/2014 0:005:15:04:57 24,175 28 05/29/2014 0:006:07:01:14 28,499 39 05/28/2014 0:006:19:03:01 32,376 58 05/27/2014 0:006:23:12:48 25,244 58 05/26/2014 0:007:02:12:23 19,811 34 05/25/2014 0:005:14:01:30 17,352 28 05/24/2014 0:004:15:21:42 15,922 27 05/23/2014 0:005:17:52:20 21,436 32 05/22/2014 0:005:23:08:53 24,955 34 05/21/2014 0:007:05:50:46 30,708 42 05/20/2014 0:007:08:50:57 28,977 40 05/19/2014 0:007:15:01:20 33,265 44 05/18/2014 0:006:22:51:42 22,398 33 05/17/2014 0:007:21:12:47 22,441 30 05/16/2014 0:005:22:23:12 20,316 30 05/15/2014 0:007:10:04:59 24,250 41 6/1/2014 7:32:35 AM | | cc_config.xml not found - using defaults 6/1/2014 7:32:35 AM | | Starting BOINC client version 7.2.33 for windows_x86_64 6/1/2014 7:32:35 AM | | log flags: file_xfer, sched_ops, task 6/1/2014 7:32:35 AM | | Libraries: libcurl/7.25.0 OpenSSL/1.0.1 zlib/1.2.6 6/1/2014 7:32:35 AM | | Data directory: C:\My Documents\BOINC 6/1/2014 7:32:35 AM | | Running under account Mark 6/1/2014 7:32:35 AM | | CUDA: NVIDIA GPU 0: GeForce GT 530 (driver version 332.21, CUDA version 6.0, compute capability 2.1, 2048MB, 1876MB available, 269 GFLOPS peak) 6/1/2014 7:32:35 AM | | OpenCL: NVIDIA GPU 0: GeForce GT 530 (driver version 332.21, device version OpenCL 1.1 CUDA, 2048MB, 1876MB available, 269 GFLOPS peak) 6/1/2014 7:32:35 AM | | Host name: M-09 6/1/2014 7:32:35 AM | | Processor: 8 GenuineIntel Intel(R) Core(TM) i7-3820 CPU @ 3.60GHz [Family 6 Model 45 Stepping 7] 6/1/2014 7:32:35 AM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 sse4_1 sse4_2 popcnt aes syscall nx lm vmx tm2 dca pbe 6/1/2014 7:32:35 AM | | OS: Microsoft Windows 7: Home Premium x64 Edition, Service Pack 1, (06.01.7601.00) 6/1/2014 7:32:35 AM | | Memory: 9.94 GB physical, 19.89 GB virtual 6/1/2014 7:32:35 AM | | Disk: 1.80 TB total, 1.72 TB free 6/1/2014 7:32:35 AM | | Local time is UTC -4 hours 6/1/2014 7:32:35 AM | | A new version of BOINC is available. (7.2.42) <a href=http://boinc.berkeley.edu/download.php>Download</a> 6/1/2014 7:32:35 AM | DistrRTgen | URL http://boinc.freerainbowtables.com/distrrtgen/; Computer ID 38436; resource share 100000 6/1/2014 7:32:35 AM | World Community Grid | URL http://www.worldcommunitygrid.org/; Computer ID 2015643; resource share 100000 6/1/2014 7:32:35 AM | World Community Grid | General prefs: from World Community Grid (last modified 07-May-2014 06:00:48) 6/1/2014 7:32:35 AM | World Community Grid | Computer location: home 6/1/2014 7:32:35 AM | | General prefs: using separate prefs for home 6/1/2014 7:32:35 AM | | Reading preferences override file 6/1/2014 7:32:35 AM | | Preferences: 6/1/2014 7:32:35 AM | | max memory usage when active: 8145.84MB 6/1/2014 7:32:35 AM | | max memory usage when idle: 8145.84MB 6/1/2014 7:32:35 AM | | max disk usage: 500.00GB 6/1/2014 7:32:35 AM | | suspend work if non-BOINC CPU load exceeds 80% 6/1/2014 7:32:35 AM | | (to change preferences, visit a project web site or select Preferences in the Manager) 6/1/2014 7:32:35 AM | | Not using a proxy 6/1/2014 7:32:36 AM | | Running CPU benchmarks 6/1/2014 7:32:36 AM | | Suspending computation - CPU benchmarks in progress 6/1/2014 7:33:08 AM | | Benchmark results: 6/1/2014 7:33:08 AM | | Number of CPUs: 8 6/1/2014 7:33:08 AM | | 3433 floating point MIPS (Whetstone) per CPU 6/1/2014 7:33:08 AM | | 11639 integer MIPS (Dhrystone) per CPU 6/1/2014 7:33:09 AM | | Suspending GPU computation - user request 6/1/2014 7:33:09 AM | World Community Grid | Restarting task MCM1_0004593_2480_1 using mcm1 version 732 in slot 7 6/1/2014 7:33:09 AM | World Community Grid | Restarting task MCM1_0004593_2827_0 using mcm1 version 732 in slot 3 6/1/2014 7:33:09 AM | World Community Grid | Restarting task MCM1_0004593_2791_0 using mcm1 version 732 in slot 5 6/1/2014 7:33:09 AM | World Community Grid | Restarting task MCM1_0004593_2890_1 using mcm1 version 732 in slot 6 6/1/2014 7:33:09 AM | World Community Grid | Restarting task MCM1_0004593_2793_0 using mcm1 version 732 in slot 0 6/1/2014 7:33:09 AM | World Community Grid | Restarting task MCM1_0004593_2882_0 using mcm1 version 732 in slot 1 6/1/2014 7:33:09 AM | World Community Grid | Restarting task MCM1_0004600_5489_1 using mcm1 version 732 in slot 2 6/1/2014 7:33:09 AM | World Community Grid | Restarting task MCM1_0004600_5123_1 using mcm1 version 732 in slot 4 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
DistrRTgen - is suspended - but when running only run on GPU!
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
6/1/2014 7:32:35 AM | | suspend work if non-BOINC CPU load exceeds 80% The Red items may explain why you do not get more than 8 CPU days 6/1/2014 7:32:35 AM | | (to change preferences, visit a project web site or select Preferences in the Manager) 6/1/2014 7:32:35 AM | | Not using a proxy 6/1/2014 7:32:36 AM | | Running CPU benchmarks 6/1/2014 7:32:36 AM | | Suspending computation - CPU benchmarks in progress 6/1/2014 7:33:08 AM | | Benchmark results: 6/1/2014 7:33:08 AM | | Number of CPUs: 8 |
||
|
|
|