| 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: Heap of Invalid units |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 16
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I have a heap of Inconclusive results gradually turning into Invalid. All of them finished at 6 or more hours due to exceeding runtime limit.
e.g. CMD2_ 0733-1W1I_ E.clustersOccur-1W98_ B.clustersOccur_ 1_ 14483_ 17159_ 2-- 614 Valid 14/08/10 03:04:42 14/08/10 07:25:30 2.87 48.1 / 68.7 CMD2_ 0733-1W1I_ E.clustersOccur-1W98_ B.clustersOccur_ 1_ 14483_ 17159_ 0-- 614 Valid 13/08/10 00:27:44 13/08/10 08:36:53 3.50 89.4 / 68.7 CMD2_ 0733-1W1I_ E.clustersOccur-1W98_ B.clustersOccur_ 1_ 14483_ 17159_ 1-- 614 Invalid 13/08/10 00:24:48 14/08/10 01:58:08 7.14 32.5 / 0.1 (To put this into perspective, they did over 4000 WUs successfully before this started happening.) Has something changed recently? |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
For a man of great detail, you obviously checked the result logs of any abnormalities. The credit is strange... 0.1, where one would expect half of the quorum which with HCMD2 is a little difficult, but at least something like half the claim for those hours.
----------------------------------------Everything here as do the stats show 20:20 A Linux 64 bit client valid result log: Result Name: CMD2_ 0682-1ZXC_ B.clustersOccur-3E0P_ B.clustersOccur_ 2_ 34638_ 35960_ 35660_ 35960_ 0-- <core_client_version>6.10.58</core_client_version> <![CDATA[ <stderr_txt> INFO: Initializing Platform. INFO: No state to restore. Start from the beginning. called boinc_finish </stderr_txt> ]]>
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
For a man of great detail, you obviously checked the result logs of any abnormalities. Indeed. Nothing abnormal at all. Usually, they are WUs which have been ended for time, so have a line such as: "Finishing early because max runtime has been exceeded.21606.534323" However, I've seen one which completed and has a totally standard log: CMD2_ 0736-1UVF_ A.clustersOccur-2QIC_ A.clustersOccur_ 0_ 2-- 614 Valid 15/08/10 16:44:13 15/08/10 19:37:12 0.88 11.9 / 13.1 CMD2_ 0736-1UVF_ A.clustersOccur-2QIC_ A.clustersOccur_ 0_ 1-- 614 Invalid 13/08/10 01:45:47 14/08/10 11:19:38 3.83 18.8 / 6.5 CMD2_ 0736-1UVF_ A.clustersOccur-2QIC_ A.clustersOccur_ 0_ 0-- 614 Valid 13/08/10 01:44:34 15/08/10 15:38:58 4.74 14.3 / 13.1 |
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges: 
|
Have they all been returned from the same machine?
----------------------------------------Also, could you please give us an idea of how much is a heap for you in this case? At least an approximative percentage.  |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
All of them finished at 6 or more hours due to exceeding runtime limit. Makes no sense as that as we all know only should be reported at 12:00 hours cut off **. Linux boxes (they, as in plural) makes it even weirder. We have one reporting lost CPU time for CEP2 but those tasks start counting new as getting a new PID on new job and something is not being passed proper to the time accumulation, but with HCMD2 nothing like that. Are these [heap] of tasks running without interruption? ** one on the RS page off the Linux quad reports running 9 hours and no cutoff logged... clean as always. CMD2_ 0697-1YVJ_ A.clustersOccur-2A9U_ A.clustersOccur_ 5_ 1-- 1292373 Valid 4-8-10 11:32:13 5-8-10 12:56:16 8.99 208.9 / 151.7 Result Name: CMD2_ 0697-1YVJ_ A.clustersOccur-2A9U_ A.clustersOccur_ 5_ 1-- <core_client_version>6.10.58</core_client_version> <![CDATA[ <stderr_txt> INFO: Initializing Platform. INFO: No state to restore. Start from the beginning. called boinc_finish </stderr_txt> ]]>
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Have they all been returned from the same machine? No. However, it's only happening to pre-P4 Intel machines. (i.e. P3 and Celeron. Athlon unaffected so far, but it's only done 4 or 5 WUs of this project recently.) Also, could you please give us an idea of how much is a heap for you in this case? At least an approximative percentage. The vast majority. Maybe 90%. e.g. Here are the most recent 10 WUs completed by one machine: CMD2_ 0746-2ZNR_ A.clustersOccur-3C5V_ A.clustersOccur_ 0_ 3432_ 8217_ 0-- davros Inconclusive 16/08/10 10:46:32 17/08/10 02:10:23 6.00 29.0 / 0.0 <-- will no doubt turn Invalid CMD2_ 0746-2FJU_ B.clustersOccur-2K4N_ A.clustersOccur_ 7_ 40611_ 42573_ 1-- davros Inconclusive 16/08/10 10:46:32 16/08/10 20:17:29 6.32 30.5 / 0.0 <-- will no doubt turn Invalid BETA_ c4cw_ beta2_ 016186666_ 0-- davros Valid 15/08/10 16:20:47 15/08/10 22:51:13 5.37 26.0 / 28.2 BETA_ c4cw_ beta2_ 006711846_ 0-- davros Valid 15/08/10 01:14:02 15/08/10 16:20:47 5.40 26.6 / 29.2 CMD2_ 0743-1TZS_ A.clustersOccur-2IAE_ F.clustersOccur_ 3_ 5711_ 5911_ 0-- davros Invalid 14/08/10 18:55:10 16/08/10 12:28:59 7.11 34.3 / 1.6 CMD2_ 0745-2CMY_ B.clustersOccur-2V17_ H.clustersOccur_ 0_ 1-- davros Invalid 14/08/10 11:37:55 16/08/10 03:48:07 6.09 29.4 / 8.3 CMD2_ 0738-1UL1_ X.clustersOccur-2R9B_ A.clustersOccur_ 6_ 51529_ 52273_ 0-- davros Valid 14/08/10 07:25:14 15/08/10 16:01:27 6.50 32.0 / 16.5 CMD2_ 0733-1W1I_ E.clustersOccur-2BTP_ A.clustersOccur_ 4_ 31053_ 32879_ 31689_ 31927_ 0-- davros Invalid 14/08/10 03:53:58 15/08/10 01:13:46 7.10 35.0 / 0.8 CMD2_ 0721-1WUU_ D.clustersOccur-2FB8_ A.clustersOccur_ 8_ 87098_ 90933_ 1-- davros Invalid 13/08/10 05:50:31 14/08/10 18:55:09 6.00 29.5 / 9.7 CMD2_ 0736-2RAK_ A.clustersOccur-2RIK_ A.clustersOccur_ 6_ 1-- davros Invalid 13/08/10 01:45:28 14/08/10 07:25:13 6.00 29.6 / 7.6 Are these [heap] of tasks running without interruption? Generally, yes. [Edit 1 times, last edit by Former Member at Aug 17, 2010 7:03:50 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
At least an approximative percentage. Recently (i.e. from current results stats page): P3/Celeron: 83% Invalid All others: 0% Invalid |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Microsoft long stopped coding for P3... Not Vista Ready! Linux, on next major release wont even run on anything below a 686. Could be time to retire.
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Could be time to retire. My tualatin P3's are faster than early P4's and use 1/3 the power. I have a 2GHz P4. It's a little bit faster on integer and a bit slower on float than a 1.4GHz P3. Really not much of an upgrade, plus hot and noisy. As this point, I'll just quote something you wrote recently: Progress is not always better nor is change for the sake of change. Happy Crunching. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Given the lack of new projects, maybe the techs could find time to look at fixing whatever they broke in August? Or just giving us the choice of letting WUs run to completion, which makes the situation much better.
|
||
|
|
|