Message boards :
Number crunching :
Unable to load 10 days worth of WUs
Message board moderation
Author | Message |
---|---|
Ghery S. Pettit Send message Joined: 7 Nov 99 Posts: 325 Credit: 28,109,066 RAC: 82 |
As of a few days ago I ceased being able to download a bunch of WUs in anticipation of the usual server serve shutdown on Tuesdays. I've got a couple of machines that don't even have a day's worth of WUs stashed and they will likely wind up idle while the weekly server work is being done. This problem appears to have started about the time I switched over to the Chicken clients and upgraded to BOINC 5.8.11. Any ideas? |
dragon1 Send message Joined: 17 Sep 05 Posts: 33 Credit: 4,438,013 RAC: 0 |
I may have the same sort of problem also...I have my cache set for 9 days but the most BOINC will provide me at any given time is about 22hrs worth of work (4 WU's). It also has developed a nasty habit of starting a WU and then downloading a new one with a much newer deadline requiring abondoning of the one just started. Then it hops all over from WU to WU starting and stopping depending on the 'latest deadline' of new WU's as they arrive. Been doing this for about a month now. |
Keck_Komputers Send message Joined: 4 Jul 99 Posts: 1575 Credit: 4,152,111 RAC: 1 |
The shortest deadline at SETI is about 4 days. Setting a queue size greater than 2 to 3 days can actually result in a smaller queue due to efforts to meet the deadlines. The best solution for project problems is a backup project. BOINC WIKI BOINCing since 2002/12/8 |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
As of a few days ago I ceased being able to download a bunch of WUs in anticipation of the usual server serve shutdown on Tuesdays. I've got a couple of machines that don't even have a day's worth of WUs stashed and they will likely wind up idle while the weekly server work is being done. When you set "connect every 10 days" you are telling BOINC that you aren't going to connect more often than that. The second connection is 20 days away. When BOINC sees work with a deadline less than 20 days away (like the typical short, 4 day deadline) it knows that it is in trouble if everything happens on schedule. If it is already in trouble (if it has work due in 4 days, and isn't supposed to connect for 10 days) it won't fetch any more work -- it won't make the deadline problem harder. Try "connect every 2 days" and see how that works for you. |
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0 |
As of a few days ago I ceased being able to download a bunch of WUs in anticipation of the usual server serve shutdown on Tuesdays. I've got a couple of machines that don't even have a day's worth of WUs stashed and they will likely wind up idle while the weekly server work is being done. Yep, reduce the connect interval to 1 day and add a cc_config.xml file with a work_request_factor of 10: <cc_config> <log_flags> </log_flags> <options> <work_request_factor>10</work_request_factor> </options> </cc_config> Joe |
Ghery S. Pettit Send message Joined: 7 Nov 99 Posts: 325 Credit: 28,109,066 RAC: 82 |
As of a few days ago I ceased being able to download a bunch of WUs in anticipation of the usual server serve shutdown on Tuesdays. I've got a couple of machines that don't even have a day's worth of WUs stashed and they will likely wind up idle while the weekly server work is being done. Thanks. That seems to have worked. Both my C2D and the multi-core Xeon system in the lab at work have down loaded more work. |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
BTW, the formula for computation deadline is Computation_deadline = report_deadline - (connect_interval + task_switch_period + 1 day). If a task cannot be completed with its resource share and the other work that is on the host in < 90% of this deadline, the task is in deadline trouble. If a task is in deadline trouble, the project that it is a part of will not download any work. The host will stop downloading work from anywhere if the sum of the work on hand is at least the queue size (i.e. a CPDN task on a single CPU host will complete the queue if a task from another project is in deadline trouble). The range for the multiplier is 1 to 10, and the maximum value for connect intervay * work fetch multiplier is 14 days. BOINC WIKI |
©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.