Forgive my being ignorant on matters networking, but can't we just tweak the
"project requested delay"? In peak times set it to a higher value and off-peak
set it to a lower value.
Einstein seems to use a 60 second delay, GPUgrid 31 seconds and Seti 11
seconds. While it won't solve all their problems it might help the server being
hit so often with requests.
13/07/2009 5:55:27 pm s...@home [sched_op_debug] Starting scheduler request
13/07/2009 5:55:27 pm s...@home sending scheduler request: To fetch work.
13/07/2009 5:55:27 pm s...@home requesting new tasks for CPU
13/07/2009 5:55:27 pm s...@home [sched_op_debug] CPU work request: 274542.11
seconds; 0 idle CPUs
13/07/2009 5:55:27 pm s...@home [sched_op_debug] CUDA work request: 0.00
seconds; 0 idle GPUs
13/07/2009 5:55:32 pm s...@home scheduler request completed: got 0 new tasks
13/07/2009 5:55:32 pm s...@home [sched_op_debug] Server version 607
13/07/2009 5:55:32 pm s...@home message from server: (Project has no jobs
available)
13/07/2009 5:55:32 pm s...@home project requested delay of 11 seconds
13/07/2009 5:55:32 pm s...@home [sched_op_debug] Deferring communication for 11
sec
13/07/2009 5:55:32 pm s...@home [sched_op_debug] Reason: requested by project
Cheers,
MarkJ
____________________________________________________________________________________
Access Yahoo!7 Mail on your mobile. Anytime. Anywhere.
Show me how: http://au.mobile.yahoo.com/mail
_______________________________________________
boinc_dev mailing list
[email protected]
http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev
To unsubscribe, visit the above URL and
(near bottom of page) enter your email address.
