| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Help Conquer Cancer
Thread: New Mxxxxx units? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 20
|
|
| Author |
|
|
coolstream
Senior Cruncher SCOTLAND Joined: Nov 8, 2005 Post Count: 475 Status: Offline Project Badges: 
|
I seem to be getting quite a few units beginning with M rather than X.
----------------------------------------One has automatically started on one machine but is reporting Output file for Mxxxxx for task Mxxxxx absent What are these M units? and does the reported message mean that there has been another configuration error at the server end?  Crunching in memory of my Mum PEGGY, cousin ROPPA and Aunt AUDREY. |
||
|
|
Mumak
Senior Cruncher Joined: Dec 7, 2012 Post Count: 477 Status: Offline Project Badges:
|
I'm too getting only those units since ~8 hours.
----------------------------------------Haven't processed any of them yet tho. Moreover, I'm recently getting many "internal server error"s when trying to report/download tasks.  [Edit 1 times, last edit by Mumak at Apr 30, 2013 8:25:37 AM] |
||
|
|
Mumak
Senior Cruncher Joined: Dec 7, 2012 Post Count: 477 Status: Offline Project Badges:
|
Forced to run an M-unit on GF 660 Ti. All processed OK. Will try on AMD too.
---------------------------------------- |
||
|
|
Mumak
Senior Cruncher Joined: Dec 7, 2012 Post Count: 477 Status: Offline Project Badges:
|
Finished a few of them on 2xHD7950 and no problem.. Of course all are PV.
---------------------------------------- |
||
|
|
coolstream
Senior Cruncher SCOTLAND Joined: Nov 8, 2005 Post Count: 475 Status: Offline Project Badges:
|
Good point to test them.
----------------------------------------I've started one too and will let it run for a while because all slots want to run other CPU units in high priority. The test M-unit appears to be check-pointing more frequently than the X-units. (8 in 20 minutes) Crunching in memory of my Mum PEGGY, cousin ROPPA and Aunt AUDREY. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Keep in mind that we're getting to a close, so the full result/batch history and delivery is audited for missing items i.e. don't be surprised to see odds and ends passing by, in fact without wanting to create high expectations, I'd not be surprised that HCC will be moved to intermittent for a bit just like where HCMD2 is, so mop up work can be passed, but, if small that work is more likely passed through WCG's own cluster.
|
||
|
|
Mumak
Senior Cruncher Joined: Dec 7, 2012 Post Count: 477 Status: Offline Project Badges:
|
Yep, we can expect at least one week of processing occasional resend units (from errors, machines that have buffered too much WUs and couldn't finish them on time, etc), then I think another week for just few resends of resends. But there's won't probably be many of them...
----------------------------------------I previously thought that these M-units are the last batch of additionally identified units, but since according to estimation there should be about 5 days of these additional ones and M-units should be expected about 1 day only, that's probably something different. |
||
|
|
coolstream
Senior Cruncher SCOTLAND Joined: Nov 8, 2005 Post Count: 475 Status: Offline Project Badges:
|
Just received a batch of GPU M-units (some resends). They have all returned 'Computation error' status.
----------------------------------------ADDITIONAL INFO: These M-units are reporting Task Mxxx exited with zero status but no 'finished' file I'll take a look at my 7000 series ATI machines and see if they have received any. Crunching in memory of my Mum PEGGY, cousin ROPPA and Aunt AUDREY. [Edit 1 times, last edit by coolstream at Apr 30, 2013 9:58:07 AM] |
||
|
|
coolstream
Senior Cruncher SCOTLAND Joined: Nov 8, 2005 Post Count: 475 Status: Offline Project Badges:
|
All GPU crunchers are receiving these ATI M-units.
----------------------------------------I am forcing some to run and they appear to be running just the same as the X-unit ones. UPDATE: all (including the resends) are running without issue and the first ones have been returned after completing in the same time as a X-unit. 2nd UPDATE: I have now stopped forcing the M-units from being run so the units that are 'high priority' should automatically start running. However, I am observing that these M-units are not allowing this and continue to run to the end. A high priority unit is only restarting after an M-unit has finished. Crunching in memory of my Mum PEGGY, cousin ROPPA and Aunt AUDREY. [Edit 3 times, last edit by coolstream at Apr 30, 2013 10:43:22 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
We need to emphasize, once more, that the first copy of a task [of all sciences except HPF2] is not locked to any platform or card. Only the additional copies are. Maybe knreed can elude on the distribution particulars to GPGPU assigned tasks [though at this time, it's FYI so near to the end]. It would be probable that if _0 were send to ATI, the _1 and on would be too.
----------------------------------------[Edit 1 times, last edit by Former Member at Apr 30, 2013 10:47:04 AM] |
||
|
|
|