| Author |
Message |
|---|
|
|
|
I just successfully completed a pps(sieve) work unit, despite the previous errors.
http://www.primegrid.com/workunit.php?wuid=253925806
In order to temporarily fix the problem, download a good number of WUs, then edit your client_state.xml (after shutting down BOINC).
Take the value for rsc_fpops_bound within the pps_sr2sieve workunits and multiply by it 100.
|
|
|
 Michael Goetz Volunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 2,965
                           
|
|
That will cure the "max time exceeded" error.
Make sure that's the problem you have before going through the trouble of editing client_state.xml. Another possible cause of computation errors on NVIDIA GPUS for these WUs is the 295 driver error.
It's VERY easy to tell which is causing the computation error:
If your run time is less than 5 seconds, it's probably the Nvidia 295 driver error.
If your run time is 1000 seconds or more, it's probably the "max time exceeded" error, and you can use the technique described by the OP.
EDIT: Thanks for the suggesting the work-around!!!
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
If you look at the unit, it was completed by an ATI card. ;)
I do have a dual configuration for NVIDIA and ATI, but my NVIDIA driver is 285.62 and it's successfully working on a genefer unit as we speak.
This fix was indeed for the universal "max time exceeded" (-177) that most people are getting.
I hope that this workaround is helpful to everyone! |
|
|
|
|
|
"In order to temporarily fix the problem, download a good number of WUs, then edit your client_state.xml (after shutting down BOINC).
Take the value for rsc_fpops_bound within the pps_sr2sieve workunits and multiply by it 100. "
I tried from a previous post and it did not work. The problem was I did not shut BOINC down. THIS IS ESSENTIAL as when continuing to run it does not normally look at this file very often. It always uses it on restart.
With a new set of WU I did it again with BOINC stopped and I have now had my first success in days.
____________
Member team AUSTRALIA
My lucky number is 9291*2^1085585+1 |
|
|
|
|
|
Yes, I probably should have bolded the "shutting down" part.
If you do not shut down the client, editing the .xml file does nothing.
Hopefully soon we won't have to do this. I'm primarily running PPSllr and Genefer WUs in the meantime, but will get more PPSsieve WUs when I have the time to edit the client_state file. |
|
|
|
|
|
yeah, thanks for the initial comment. I have got many done now.
I also got one due 15th that has run without any editing, perhaps it is all OK now.
____________
Member team AUSTRALIA
My lucky number is 9291*2^1085585+1 |
|
|
|
|
|
It's possible that doing the initial editing of the client_state.xml may have assisted in automatically lengthening the ETA?
After finally being able to fix my computer, I'm going to try getting another batch of them. Maybe it was resolved by an admin, or this "fix" took care of it.
Either way, if I hit another error, I'm editing the .xml file. :P |
|
|
