Message boards : Number crunching : Problems with Rosetta version 5.82
Previous · 1 · 2 · 3 · 4
Author | Message |
---|---|
Clinton H Send message Joined: 21 Apr 08 Posts: 1 Credit: 4,620,233 RAC: 17,912 ![]() |
Resently (almost the past two weeks actually) all of the Rosetta Mini tasks have failed to start up or complete. They stay at 0:00:00 CPU time and do not advance or fail. If anyone has a solution as to how to not recieve or how to ignore these computations or a fix for the Rosetta Mini application please respond. |
![]() ![]() Send message Joined: 2 Jul 06 Posts: 2842 Credit: 2,020,043 RAC: 0 |
what happen to this ???https://boinc.bakerlab.org/rosetta/results.php?hostid=848476 |
Mod.Sense Volunteer moderator Send message Joined: 22 Aug 06 Posts: 4018 Credit: 0 RAC: 0 |
what happen to this ???https://boinc.bakerlab.org/rosetta/results.php?hostid=848476 That machine appears to have downloaded a bunch of work, assuming the default 3hr runtime, and was running under a 6hr runtime preference. Further, the tasks that were complete were restarted many times. This indicates BOINC was ended, or other task preempted Rosetta and tasks are not being retained in memory when suspended. The rest of the tasks were not reported back before their deadline. ...have I missed the question? Rosetta Moderator: Mod.Sense |
![]() ![]() Send message Joined: 2 Jul 06 Posts: 2842 Credit: 2,020,043 RAC: 0 |
what happen to this ???https://boinc.bakerlab.org/rosetta/results.php?hostid=848476 im trying to help out someone else machine.....whats the best thing to do to fix it? |
Mod.Sense Volunteer moderator Send message Joined: 22 Aug 06 Posts: 4018 Credit: 0 RAC: 0 |
im trying to help out someone else machine.....whats the best thing to do to fix it? I would do the following: Set computing preferences to leave applications in memory while preempted. And the networking preferences to connect every .1 days with zero extra days of work (this will prevent you from getting more then you can crunch). Keep in mind these are possibly specific to the venue of the machine. Then abort all of the tasks that have passed their deadline. Then update to the project. That should result in your getting some new work downloaded. Check the estimated time to completion. Crunch and report a few work units until the estimated time to completion of a new task is roughly the 6hrs of your current Rosetta runtime preference. Once estimated runtime is near your runtime preference, then adjust the network settings to cache more work or get extra work on each request as you prefer. Appears this machine is not likely on all the time. BOINC, when first installed, doesn't know what % of the time the machine tends to be running BOINC. As you crunch and report a few more tasks, BOINC will start to realize that a machine that is only powered on 20% of the time only needs 20% of the work to keep it busy for a day (or whatever your friend's power on % happens to be). So, it will learn to avoid missing the deadlines, by not requesting too much work. Rosetta Moderator: Mod.Sense |
![]() ![]() Send message Joined: 2 Jul 06 Posts: 2842 Credit: 2,020,043 RAC: 0 |
|
Mod.Sense Volunteer moderator Send message Joined: 22 Aug 06 Posts: 4018 Credit: 0 RAC: 0 |
https://boinc.bakerlab.org/rosetta/result.php?resultid=189472246 Is there a type-o somewhere? That machine doesn't show you as the owner and the task completed normally and is not v5.82 (the topic of this thread) Rosetta Moderator: Mod.Sense |
Message boards :
Number crunching :
Problems with Rosetta version 5.82
©2023 University of Washington
https://www.bakerlab.org