I am on about a gap in my results,My results before are credited ok,
then a run of results that are still pending,
then the following results are all ok and credited
I am on about a gap in my results,My results before are credited ok,
then a run of results that are still pending,
then the following results are all ok and credited
Have you checked to see if there is a user who has taken a bunch of jobs but not returned any work, such as this: http://einsteinathome.org/host/254029
This users has taken over 20 jobs & returned none as of (~15:00 UTC 24 Jan 2006):
This seems to be an increasing phenomenon; some users take many jobs & then do not return results causing our Pending amounts to rise, but no credit granted. Also, it seems that the scheduling program often groups the same users together, because I find I often have the same user assigned to me to form many quorums.
PS: I do not mean to pick on the user I have listed above, but he just happened to be one of many I get.
I'm not sure what you consider to be "alot" of Pending credit, but I rather suspect that it's all relative. My own current Pending credit sits at roughly 6,400 credits but my RAC is also at roughly 3,600 - so only just less than 2 days worth. No complaints here - at a much faster rate of production, I expect to have a pile in pending.
BTW, my thanks to the E@H/BOINC project admins for keeping the project stocked with work so that I can continue to crunch! :)
"Chance is irrelevant. We will succeed."
- Seven of Nine
This seems to be an increasing phenomenon; some users take many jobs & then do not return results causing our Pending amounts to rise, but no credit granted. Also, it seems that the scheduling program often groups the same users together, because I find I often have the same user assigned to me to form many quorums.
Best way to handle that: Never ever check your pendings :-)
I'm not sure what you consider to be "alot" of Pending credit, but I rather suspect that it's all relative.
Yes, I meant our relative amounts. My point, or observation, was that there were users who download ('reserved') large blocks of work, relative to their output & many of them do not seem to finish most, and often any of their work. Thus, the quorum is not obtained until the time limit expires & a new cruncher can finish the job.
I noticed it recently when my rate dropped & I assumed it was due to a crash of the Einstein program on 1 of only 2 machines I have crunching. It wasn't & Einstein has never crashed, but it caused me to find out why there were no work units being completed.
I'm new to BOINC & Einsteain, so I was wondering if it is a flaw of the assigner program or is the user's computer flailing away requesting more work & yet never finishing it?
Pendingresults
)
Uhm... Whatever you do please do not blame Michael. :)
I am on about a gap in my
)
I am on about a gap in my results,My results before are credited ok,
then a run of results that are still pending,
then the following results are all ok and credited
Yes, so? You do know how
)
Yes, so?
You do know how Credits are calculated, claimed and granted under any BOINC project, don't you?
If you don't...
Read in the BOINC-Wiki:
Credit
Claimed Credit
Granted Credit
RE: I am on about a gap in
)
Have you checked to see if there is a user who has taken a bunch of jobs but not returned any work, such as this:
http://einsteinathome.org/host/254029
This users has taken over 20 jobs & returned none as of (~15:00 UTC 24 Jan 2006):
http://einsteinathome.org/host/254029/tasks
This seems to be an increasing phenomenon; some users take many jobs & then do not return results causing our Pending amounts to rise, but no credit granted. Also, it seems that the scheduling program often groups the same users together, because I find I often have the same user assigned to me to form many quorums.
PS: I do not mean to pick on the user I have listed above, but he just happened to be one of many I get.
I'm not sure what you
)
I'm not sure what you consider to be "alot" of Pending credit, but I rather suspect that it's all relative. My own current Pending credit sits at roughly 6,400 credits but my RAC is also at roughly 3,600 - so only just less than 2 days worth. No complaints here - at a much faster rate of production, I expect to have a pile in pending.
BTW, my thanks to the E@H/BOINC project admins for keeping the project stocked with work so that I can continue to crunch! :)
"Chance is irrelevant. We will succeed."
- Seven of Nine
RE: This seems to be an
)
Best way to handle that: Never ever check your pendings :-)
My most recent personal peak
)
My most recent personal peak pending credit was 44,000. down to about 20,000 now.
RE: I'm not sure what you
)
Yes, I meant our relative amounts. My point, or observation, was that there were users who download ('reserved') large blocks of work, relative to their output & many of them do not seem to finish most, and often any of their work. Thus, the quorum is not obtained until the time limit expires & a new cruncher can finish the job.
I noticed it recently when my rate dropped & I assumed it was due to a crash of the Einstein program on 1 of only 2 machines I have crunching. It wasn't & Einstein has never crashed, but it caused me to find out why there were no work units being completed.
I'm new to BOINC & Einsteain, so I was wondering if it is a flaw of the assigner program or is the user's computer flailing away requesting more work & yet never finishing it?
Here are some examples & users I frequently seem to get assigned with to a quorum:
http://einsteinathome.org/host/53203/tasks
http://einsteinathome.org/host/16908/tasks
http://einsteinathome.org/host/513257/tasks
mark: how're you getting
)
mark: how're you getting such high levels of RAC/comp? I've got an dualcore athlon @2.4gig but am only getting ~1/3rd the RAC of your opteron system.
Dan I have two dual core
)
Dan
I have two dual core opteron cpus for a total of 4 cpus. Also running the boinc manager from here Alpha Main Page