| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Help Cure Muscular Dystrophy
Thread: Very Long Work Units |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 30
|
|
| Author |
|
|
davidhobbs
Senior Cruncher England Joined: Dec 30, 2004 Post Count: 152 Status: Offline Project Badges: 
|
Just an observation:
I currently have two work units in progress that have reached approximately 60% after 44 hours and 48 hours respectively. Admittedly they are running on fairly slow machines (600MHz) but I think these are the longest times I seen on this project. David. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
There was a backroom post that more segments and seeds were packed into the FA@H and HPF2 jobs to help buffer the additional server load due the large influx of new members and the fail-over of SETI clients to WCG (SETI is of air due server crash likely thru Wednesday). Latter would though only impact the BOINC server loads.
----------------------------------------Presume the HCMD job is clocking CPU time in the Taskmanager but more than normal. Can you confirm? cheers
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at May 6, 2007 4:04:34 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I am also observing a change in the HCMD workunits which is not good.
I am observing the following 1) The CPU throttle is not working until after a checkpoint occurs-the wcgridHCMD.exe consuming 100% of the CPU until then. 2) The checkpoints appear to be too far apart, for my particular workunit it seems to be about 90minutes for a 4% checkpoint. Therefore, if I don't leave my computer on for more than 90minutes, then all the data is lost. Plus wcgridHCMD is not obeying the throttle during that 90minutes. I imagine this is not the intended behavior to have checkpoints this far apart. This is an average loss of 45min everytime the process needs to be restarted. If this continues on the next workunit, I'm going to drop HCMD from my project list. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I am also observing a change in the HCMD workunits which is not good. I am observing the following 1) The CPU throttle is not working until after a checkpoint occurs-the wcgridHCMD.exe consuming 100% of the CPU until then. 2) The checkpoints appear to be too far apart, for my particular workunit it seems to be about 90minutes for a 4% checkpoint. Therefore, if I don't leave my computer on for more than 90minutes, then all the data is lost. Plus wcgridHCMD is not obeying the throttle during that 90minutes. I imagine this is not the intended behavior to have checkpoints this far apart. This is an average loss of 45min everytime the process needs to be restarted. If this continues on the next workunit, I'm going to drop HCMD from my project list. 3) One additional note: Snooze button does not work (not sure if it works after the first checkpoint). One of my key points of a distributed program is that it should allow users to retain control over the computer. This definitely is not working as it should. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
20 hour --> 6%
I've never had anything like this. Should I abort? . |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello oliviathefirst,
Looking at http://www.worldcommunitygrid.org/forums/wcg/viewthread?thread=12910 we seem to be getting some long but variable work units. Nobody has reported any that are actually stuck, so I suggest sticking with them. See how long it actually takes to finish your work unit. Lawrence |
|||
|
||||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello oliviathefirst, I'll do that Lawrence . Thanks. I've a very patient person.Nobody has reported any that are actually stuck, so I suggest sticking with them. See how long it actually takes to finish your work unit. Lawrence It's moved up to 7% (21 hours) so it is progressing. I leave my machine on 24 hours so I may be surprised in the morning. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Protein-protein interactions have a LOT of calculating to do
 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Protein-protein interactions have a LOT of calculating to do The amount of calculations is not my point. I could care less if the WU takes 400hrs to complete. My point is 1) the time between checkpointing is too long. This is wasteful from an efficiency standpoint, but it still doesn't really bother me except that: 2)The process does not honor the CPU throttle setting or Snooze command until a checkpoint occurs. Someone just please admit/agree that there is a problem instead of defending a bug in the program. [Edit 1 times, last edit by Former Member at May 9, 2007 4:28:12 AM] |
||
|
|
|