| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Help Cure Muscular Dystrophy - Phase 2 Forum
Thread: Really long running unit [RESOLVED] |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 6
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I have an HCMD2 unit that has been running for 22+ hours, and is at 55% with ~16 hours to completion.
----------------------------------------Abort or crunch on? The WU is CMD2_1749-2ZMC_A.clustersOccur-3CTZ_A.clustersOccur_12, and I can see that my buddy sent it in after a 6 hour crunch. [Edit 1 times, last edit by Former Member at May 13, 2011 7:04:23 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I ended up aborting the unit, because the % complete had not moved at all for several hours.
After aborted, BOINC hung for a bit, and gave the following in the message dialog: [error] garbage_collect(); still have active task for acked result CMD2_1749-2ZMC_A.clustersOccur-3CTZ_A.clustersOccur_12_1; state 5 [error] garbage_collect(); still have active task for acked result CMD2_1749-2ZMC_A.clustersOccur-3CTZ_A.clustersOccur_12_1; state 6 |
||
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3315 Status: Offline Project Badges: 
|
Something could have been hogging the CPU.
----------------------------------------You were probably watching wall clock time. If an HCMD2 WU is below 60% and 6 hours(CPU time) have come and gone then the WU will end. If after 6 hours(CPU time) the WU has passed 60% then it will continue until 12 hours (CPU time) it will end regardless of the percentage done.  - AMD Ryzen 5 1600AF 6C/12T 3.2 GHz - 85W - AMD Ryzen 5 2500U 4C/8T 2.0 GHz - 28W - AMD Ryzen 7 7730U 8C/16T 3.0 GHz |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I ended up aborting the unit, because the % complete had not moved at all for several hours. After aborted, BOINC hung for a bit, and gave the following in the message dialog: [error] garbage_collect(); still have active task for acked result CMD2_1749-2ZMC_A.clustersOccur-3CTZ_A.clustersOccur_12_1; state 5 [error] garbage_collect(); still have active task for acked result CMD2_1749-2ZMC_A.clustersOccur-3CTZ_A.clustersOccur_12_1; state 6 hi beat me500, As per Falconet, it's probable something else that was eating the CPU time, whilst you allowed BOINC to continue. Best is if in doubt, to open the BOINC manager in advanced view, click the task is question and then hit the properties button on left. That will open a window which tells the amount of Wallclock/Elapsed time and the real allocated CPU time. For a HCMD2 task latter will never exceed 12 hours. To see any rampant process, a bot process maybe, hit the Ctrl-Shft-Esc keys in that order simultaneous to open the Task Manager, hit the Show all user processes button left bottom, then sort them on CPU % and or CPU time (column can be added), to see what's running. At times I've seen a svchost.exe go mad, there can be a dozen at the same time. If so, kill it, at least, I've never found this to have repercussions on crunching. Let us know. --//-- PS, the garbage collect is basically a clean up of control files when a task has ended inordinately. Presume you aborted the task via the BOINC Manager? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Thanks Falconet and SekeRob.
I actually had checked the task manager and the correct amount of units were the only things consuming CPU, and all were using the same amount. All other tasks were completing as expected. I had not checked the preferences, so I could not fully confirm what was going on. I did abort the unit, and it has not reproduced, so I think we can let this one go. Your comment about the garbage collector sounds spot on. It was a manual abort through the manager. Thanks again for the replies. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Adding one more interesting tidbit:
When I checked my Windows task manager today, I had one HCMD2 process running, even though BOINC said I did not! The process was not using CPU resources, and only ~3MB memory. I had the correct full amount of other WU's running. Strange... I killed the process. |
||
|
|
|