BOINC v6 to v7 when on Anonymous platform

Message boards : Number crunching : BOINC v6 to v7 when on Anonymous platform
Message board moderation

To post messages, you must log in.

AuthorMessage
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13736
Credit: 208,696,464
RAC: 304
Australia
Message 1381434 - Posted: 15 Jun 2013, 8:20:10 UTC
Last modified: 15 Jun 2013, 8:49:36 UTC

***Don't do it unless you run down your cache first***


I read the release note about the 197 (0xc5) EXIT_TIME_LIMIT_EXCEEDED problem, so i went for option 2 and reset the project after upgrading BOINC.

Unfortunately the Scheduler has no idea of what goes where when it comes to resends-
On one system for some reason most of the CPU work expired (pretty much all that which didn't get allocated to the GPU). I now have a daily quota of 1 WU, which i've reached.


On the other system CPU work was allocated to the GPU (only a problem with the VLARs). GPU work as been allocated to the CPU.
EDIT- exited BOINC on this system & restarted- CPU work now allocated to CPU, GPU work now allocated to GPU. Did the same on the other system- it requested GPU & CPU work, and got work for the GPU. Very next request- get hit with the 1 WU daily limit again.

Also just looked at my Computers, Details, Application details.
On both systems there are 0 consecutive valid tasks for the CPU. 1,763 for the GPU on one system, 6 on the other (the one that also has the 1 WU daily quota).



Don't know if it's relevant or not- since the upgrade the Project, Properties window no longer has a DCF. The system that didn't expire most of the work still has a DCF there.
I guess it'll all just add to the already plummeting RAC.
Hopefully it'll sort itself out in a few days.
Grant
Darwin NT
ID: 1381434 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13736
Credit: 208,696,464
RAC: 304
Australia
Message 1381786 - Posted: 16 Jun 2013, 9:29:58 UTC - in response to Message 1381434.  

Hopefully it'll sort itself out in a few days.

Pleasantly surprised, it had pretty much untangled itself after about 12 hours or so. Because of the long CPU crunching times on the system that had all the WUs time out & ended up with the 1 WU/day limit it still occasionally bumps up against it's present limit, but as it completes work & it's validated that limit keeps rising.
The limit is presently 37WU/day, but luckily most of the ones it's got at the moment are VLARS, so at over 4hrs/WU i've pretty much got a full cache even with less than the maximum server side limit.
Grant
Darwin NT
ID: 1381786 · Report as offensive
Josef W. Segur
Volunteer developer
Volunteer tester

Send message
Joined: 30 Oct 99
Posts: 4504
Credit: 1,414,761
RAC: 0
United States
Message 1381844 - Posted: 16 Jun 2013, 16:01:25 UTC - in response to Message 1381434.  

...and reset the project after upgrading BOINC.
...
Don't know if it's relevant or not- since the upgrade the Project, Properties window no longer has a DCF. The system that didn't expire most of the work still has a DCF there.
...

If a host DCF for a project is exactly 1.0 then it isn't shown by BOINC 7.0.54 or later. A change in BOINC version sets it to 1.0, and with the dont_use_dcf line in Scheduler replies to BOINC 7 clients it doesn't change.
                                                                   Joe
ID: 1381844 · Report as offensive

Message boards : Number crunching : BOINC v6 to v7 when on Anonymous platform


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.