| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Microbiome Immunity Project
Thread: WU keeps resetting |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 5
|
|
| Author |
|
|
tvdsluis
Cruncher Joined: Apr 12, 2007 Post Count: 26 Status: Offline Project Badges: 
|
This is a problem i have not seen yet.
I have a few work units, even new ones, that keep resetting after an exit of the boinc manager. I have 4 cores running. These WU's get to lets say 46% after 3 hours running. Then i close boinc because i shut my computer since i dont leave it running when im not at home. Then they all start at 0% again. I know you always lose some % between stop and start but this sounds a bit high. Are there now WU's that only complete when you process them to 100% at one sitting? Seems to waste a lot of cpu this way. Any help is appreciated. btw no strange things or errors in the log file. |
||
|
|
ca05065
Senior Cruncher Joined: Dec 4, 2007 Post Count: 328 Status: Offline Project Badges: 
|
MIP work units can only checkpoint at the end of a structure. The number of structures in a work unit can be seen after completion in the results log within the results status section of the website. It can also be seen in the stderr.txt file in the appropriate slots folder as the nstruct parameter while the work unit is running.
If you are using Windows you can use Boinctasks as a BOINC manager which shows the time since the last checkpoint for each running work unit. Instead of shutting down your PC, there is the option to hibernate the PC when it will write the current state of the running PC to disk and resume from there on the next start and no time will be lost. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The CPU time since last checkpoint can be seen by going to the advanced view of BOINC Manager, selecting the task, then hitting the Properties button.
The checkpoint interval varies both within and between the different sub-projects. You might want to check that you haven't set "Request tasks to checkpoint at most every" in Options -> Computing preferences... Computing tab to be very long. I agree with trying hibernation. You will find your machine starts up much quicker, too. I can't remember if you need to set Leave non-GPU tasks in memory while suspended, but I always have that set anyway. You'll find it in the Disk and memory tab of Computing preferences. If none of this works, then I suggest you post details of specific WUs causing the problems. I don't run all sub-projects, and others may know if you've chosen to run something that has extremely long checkpoint intervals, or if something else might be amiss. |
||
|
|
tvdsluis
Cruncher Joined: Apr 12, 2007 Post Count: 26 Status: Offline Project Badges:
|
Thanks for your anwers.
I looked at the nstruct parameter in the stderr.txt file and found indeed the value 1 for the WU's that kept resetting. Also values 6 11 and 12, so the "mystery" about the resets is solved. I'm now also using hibernate during the day and only shutdown for longer periods. Thanks! |
||
|
|
Speedy51
Veteran Cruncher New Zealand Joined: Nov 4, 2005 Post Count: 1326 Status: Offline Project Badges:
|
I am pleased you have managed to sort your issue out. When I used hibernation and number of years ago on a different project I found it to be annoying at times. Reason being is if the mouse got touched just slightly the computer would start itself this was annoying because my computer is in my room so sometimes it would wake me up at night. Other than that it is a great invention
---------------------------------------- |
||
|
|
|