| 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: Elapsed time weirdness [RESOLVED] |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 34
|
|
| Author |
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
There's a BM compile in circulation for windows only, 6.10.76, that shows CPU and Elapsed side by side, but now exclusive rely on BOINCTasks (latest v 0.63) to monitor. That one shows Elapsed time and in brackets CPU time and stores that also in the completed tasks history. All clients/tasks in 1 view and sortable.
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
eviltoad
Senior Cruncher Australia Joined: Nov 5, 2005 Post Count: 190 Status: Offline Project Badges: 
|
Ah, thanks; that's good to know.
----------------------------------------Just as a matter of interest: when the WU has been completed, the Elapsed Time counter in Advanced View changes to the actual CPU time, thus giving the impression of lost credit.  |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
I'll take a mental note of the observation. To understand for myself, the Ready to Report is then also showing with the CPU time rather than the wallclock time it took. I'm running 6.10.17 too [on Lucid Lynx], so will take extra care in observing. I've suspended networking for the moment so control the next uploads and reportings.
----------------------------------------Thanks for sharing. PS: Elapsed in the simple view is still CPU time to add to the confusion :(. If you open a second BM and switch that to simple view and overlay that on the Advanced View BM you can watch both.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
OK, first part of the trial to replicate with 6.10.17 on Lucid Lynx:
----------------------------------------1. Set the client to network suspended. 2. Two CEP2 tasks finished, upon which the percent jumped to 100%, but the Wallclock/Elapsed time remained unchanged, the status showing as "Uploading". The Task Properties screen confirmed this, showing an about 20m minutes lesser CPU time. 3. Set the client to network always 4. Uploading started and elapsed times did not change all through the uploading. And then I missed to watch what happened when it got to the Ready to Report stage, so suspended network again to observe this on the next release. ttyl
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
I'm sorry, but have been unable to replicate... even with the state of Ready To Report, the Elapsed time continued to show until clearing. Hoping someone else can reproduce.
----------------------------------------And almost on command, the somewhat longer tasks start appearing. Now have one over 30% progress at 3:30 CPU hours and continuing to checkpoint about every 30 minutes. That would make it already 45 minutes more than the previous longest here.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
eviltoad
Senior Cruncher Australia Joined: Nov 5, 2005 Post Count: 190 Status: Offline Project Badges:
|
Curious! I'll try to replicate it again for myself, with screenshots tomorrow. I'm reverted to Karmic after the online Lucid upgrade bombed and the USB and CD installs failed as well.
----------------------------------------Then I downloaded the newer BOINC client... |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
I've discovered something more than weird in Lucid. Do a 'top' command and look at the CPU time used by the science processes. Compare that to the CPU time in BOINC Manager. Have verified this with the System Monitor (SM) tool. Both show only a small part, identical value of the real time that BOINC reports, though the processes run uninterrupted... e.g. one is now running 3:10 hours and Top/SM show 28 minutes. There are no pause entries in the BM message log. Can you tell what you see in Karmic?
----------------------------------------thx
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
At least Jaunty shows the same behaviour. This has something to do with restarting after a checkpoint/part-job being finished. From my lucid:
boinc 4557 99.5 2.6 299812 54212 ? RN 08:19 1:55 ../../projects/www.worldcommunitygrid.org/wcgrid_cep2_qchem_6.19_i686-pc-linux-gnu A.19.C15H9N3O.68.bp86.svp.n.bp86.svp.n.sp.in ./ A.19.C15H9N3O.68.bp86.svp.n.bp86.svp.n.sp 5080486600156925579 54439 0 after checkpointing: boinc 4563 91.2 1.9 289656 38944 ? RN 08:21 0:06 ../../projects/www.worldcommunitygrid.org/wcgrid_cep2_qchem_6.19_i686-pc-linux-gnu A.19.C15H9N3O.68.bp86.svp.n.b3lyp.svp.n.sp.in ./ A.19.C15H9N3O.68.bp86.svp.n.b3lyp.svp.n.sp 3850045900478670044 43567 0 look at the PID. The time shown by top matches rather closely the time since last checkpointing (could you recheck that?) |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Think you're right on the money with that. Running CEP2 only on that client, times seen never seem to exceed the duration of a job within a task. If not heard from me again on this, I've confirmed.
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
petehardy
Senior Cruncher USA Joined: May 4, 2007 Post Count: 318 Status: Offline Project Badges: 
|
On my VMWare virtual machine the initial value of "Time Left" is about right ~3hrs, but this value increases throughout the job, until it ends at about 2x the initial value.
----------------------------------------As if the logic is somehow reversed.  "Patience is a virtue", I can't wait to learn it! |
||
|
|
|