Not getting new work, and the "Note:not requesting new work" message
11 Oct 2006 21:36:10 UTC
Topic 191938
(moderation:
Could anyone explain what's going on with this message I've gotten today -- perhaps in previous days too. I haven't gotten any new work from Einstein for quite a while, and am getting this UNWANTED message as I try to update:
10/11/2006 5:21:45 PM|Einstein@Home|Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
10/11/2006 5:21:45 PM|Einstein@Home|Reason: Requested by user
10/11/2006 5:21:45 PM|Einstein@Home|Note: not requesting new work or reporting results
10/11/2006 5:21:50 PM|Einstein@Home|Scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi succeeded
I could send the whole logfile, but nothing else seems amiss. None of my preferences that I can access give any hint as to what might be triggering this. Maybe it's epiphenomenal -- not related to the crux of the problem at all.
Thanks!
m
Not getting new work, and the "Note:not requesting new work" mes
)
Most likely your host has done extra work for Einstein lately and has to repay it to the other projects. It could also be that extra work is needed for another project now. If you just let it be it will get back to Einstein when it is ready.
BOINC WIKI
BOINCing since 2002/12/8
RE: Could anyone explain
)
I'm getting the say message you are getting("not requesting new work or reporting results"). Oh well, there's other projects to crunch.
RE: I'm getting the say
)
It’s most likely those other projects—or, more precisely, their long-term debt and resource-sharing situation as concerns E@h—that’s causing the behaviour. Let BOINC ‘balance the books’ as it deems necessary and you’ll see more E@h work soon.
I crunch for another project that put several of my hosts into EDF mode (Earliest Deadline First) for most of July and August, due to grossly underestimated WU sizes. The systems all stopped requesting work from that project for several weeks afterward, but recently, one by one, they’ve been resuming on their own (very cautiously, I might add, having learned from experience that those WUs can be ‘bigger than they look’, multiplying the project’s estimates by a large factor). The moral of the story is that the current BOINC scheduling routine seems to do a pretty good job of honouring the resource shares you assign while adapting to circumstances, even in the face of extreme situations, if you refrain from micromanaging it or ‘second-guessing’ its decisions.
P.S. I’ll admit to hitting the Update button now and then, to make sure the systems were still attached, but I’m sure it wasn’t anything I did that caused the work requests to resume.
Yeah, whats going on. I have
)
Yeah, whats going on. I have not gotten any new work for over 5 days.
Aha!! Daylight at last. Went
)
Aha!! Daylight at last. Went away for 3 months and when I returned seti, LHC & einstein refused further work. Climateprediction was only one that kept on. That was 3 months ago. Have tried to get more work for others (update etc) but to no avail.
climatpred is massive. Still got 25% & 38% to by end of year.
So its BOINC priotitising!!!
RE: Aha!! Daylight at last.
)
Suspend climateprediction and you will get SETI, Einstein and others. This is what I do when I want to do some work with a nearer deadline than climateprediction.
Tullio
I too had BOINC stop getting
)
I too had BOINC stop getting Seti and Einstein WUs. After their last WUs were reported, no other WUs were "requested" because according to BOINC:
"Suspending work fetch because computer is overcommitted."
Using earliest-deadline-first scheduling because computer is overcommitted."
For Climate Prediction, it has 63% to go before completion. With an estimated 1,087 hours to go and a deadline of 1.18.2007, it appears BOINC wants to commit all its time to the complete Climate Prediction WU.
JimPas