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: CPU Time vs Elapsed Time |
No member browsing this thread |
Thread Status: Active Total posts in this thread: 13
|
Author |
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
@jonnieb-uk (wow, lots of UK users on this forum)
----------------------------------------7.2.0 compiled from source using gcc bdver1 and O2 optimizations. Running on Gentoo Linux (64 bit) with 32xCores and 64GB of Ram. @lawrencehardin, I don't believe it is running/restarting from checkpoint as the server has not restarted since end-Jan and boinc/boinc_mgr has not restarted since I moved the working directory from the SSD to the Ramdrive. I'll restart the computer once I get home today and see if the problem still exists. If it does, I'll try and stopwatch a WU to see if elapsed time matches wall-clock time. [Edit 1 times, last edit by Former Member at Feb 10, 2014 3:52:21 PM] |
||
|
Yarensc
Advanced Cruncher USA Joined: Sep 24, 2011 Post Count: 134 Status: Offline Project Badges: |
I know this is a bit old at this point, but if you were still wondering, the heavy I/O periods are from checkpointing (CEP2 has quite large files) as there isn't any I/O during the actual computation. It will revert back to the checkpoint whenever the task pauses for high CPU usage if you don't check "Leave applications in memory while tasks suspended" (LAIM)
All projects on WCG run single threaded jobs, so there's no way that could have made your cpu time higher than the elapsed, as far as I know it shouldn't be possible. |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I did check Leave Applications in Memory while suspended, and am running out of a 10GB RAMDrive to stop the write bottleneck.
To try and alleviate the issue, I've dropped back down to only running 1xCEP2 at a time, and have tried a reboot and different Linux Kernels but still have the same issue. MCM also has CPU Time higher than elapsed time. I agree it shouldn't be possible, but am wondering if it is because of the Bus/CPU overclock or All-Core Turbo mode. More info here: https://secure.worldcommunitygrid.org/forums/wcg/viewthread_thread,36394 |
||
|
|