Posts by D

1) Message boards : Number crunching : Problems with web site (Message 67385)
Posted 27 Aug 2010 by D
Post:


rosetta@home Scheduler request completed: got 0 new tasks


THanks

D


You are not alone, same problem here. 2 computers, rosetta says no work available while other projects seem to work just fine!


Looks like IMA back to 100%. All machines and cores have tasks to crunch. u?
2) Message boards : Number crunching : Problems with web site (Message 67372)
Posted 26 Aug 2010 by D
Post:
PS. On my main rig I noticed the last remaining task has been running for 47 hours and reports 83 hours to completion and the time to completion is going UP.


Click on the properties tab and look at the difference between elapsed and cpu time. If there is a large difference then you are going nowhere very slowly. In these situations I close down and restart boinc which usually does the trick.



yup, CPU time is 1.25 hours vs elapsed of nearly 50 hours. I'm trying to pull the other 2 units which appear to be unaffected out of the machine and then reboot.

Everything else seems to be normal. Thanks for the fast reply!

D


Your problem may be caused by checkpointing problems on that work unit. Sometimes a work unit won't create a checkpoint/save progress so you do 4 hours work, switch to a new task, switch back again and find you are on 4 hours elapsed with 0 hours cpu time as it has gone back to beginning. Repeat that enough times and you end up with your situation of 50 hours elapsed but minimal cpu time.

If the problem persists after you reboot then two possible solutions are to either turn on the "leave suspended tasks in memory" option or abort the task.




Update: 345pm EST

Boinc software purged bad work unit on it's own. Que is now cleared.

Have enabled 'leave suspended tasks in memory' and rebooted and reset client. I've waited about 30 minutes now to see if I get new tasks or downloads.

Disk usage is only ONE MB as I have no data to crunch.

All work units have now been completed. I have no tasks and my requests to update result in these messages:

8/26/2010 3:36:49 PM rosetta@home Resetting project
8/26/2010 3:36:50 PM rosetta@home Sending scheduler request: To fetch work.
8/26/2010 3:36:50 PM rosetta@home Requesting new tasks for GPU
8/26/2010 3:36:53 PM Re-reading cc_config.xml
8/26/2010 3:36:53 PM Re-read config file
8/26/2010 3:36:53 PM log flags: file_xfer, sched_ops, task
8/26/2010 3:36:55 PM rosetta@home Scheduler request completed: got 0 new tasks
8/26/2010 3:37:02 PM No general preferences found - using BOINC defaults
8/26/2010 3:37:02 PM Reading preferences override file
8/26/2010 3:37:02 PM Preferences:
8/26/2010 3:37:02 PM max memory usage when active: 2151.99MB
8/26/2010 3:37:02 PM max memory usage when idle: 2483.07MB
8/26/2010 3:37:02 PM max disk usage: 10.00GB
8/26/2010 3:37:02 PM (to change preferences, visit the web site of an attached project, or select Preferences in the Manager)
8/26/2010 3:41:00 PM rosetta@home Sending scheduler request: To fetch work.
8/26/2010 3:41:00 PM rosetta@home Requesting new tasks for GPU
8/26/2010 3:41:02 PM rosetta@home Scheduler request completed: got 0 new tasks
8/26/2010 3:45:07 PM rosetta@home Sending scheduler request: To fetch work.
8/26/2010 3:45:07 PM rosetta@home Requesting new tasks for GPU
8/26/2010 3:45:09 PM rosetta@home Scheduler request completed: got 0 new tasks

PS - thanks for having a look. I'm checking my client status and board every 15-30 minutes or so.

THanks

D
3) Message boards : Number crunching : Problems with web site (Message 67365)
Posted 26 Aug 2010 by D
Post:
PS. On my main rig I noticed the last remaining task has been running for 47 hours and reports 83 hours to completion and the time to completion is going UP.


Click on the properties tab and look at the difference between elapsed and cpu time. If there is a large difference then you are going nowhere very slowly. In these situations I close down and restart boinc which usually does the trick.



yup, CPU time is 1.25 hours vs elapsed of nearly 50 hours. I'm trying to pull the other 2 units which appear to be unaffected out of the machine and then reboot.

Everything else seems to be normal. Thanks for the fast reply!

D
4) Message boards : Number crunching : Problems with web site (Message 67361)
Posted 26 Aug 2010 by D
Post:
I'm having some sort of problem as of 8/25 around midnight EST.

All of my computers have been unable to get new tasks. My solo and dual core machines have been out of a job and twiddling thumbs for about 6 hours now and my main machine is down to 50% and will be out of work in about 2 hours.

I've clicked 'update' and 'reset', rebooted , lit incense and laid silicon offerings (jar of sand and old CYRIX processor) in front of the machines - all to no avail. They no hungry.

Seriously, it just appears that scheduler has no new tasks.

I checked the server status and it's all green. Internet connectivity is not an issue.

I want to rule out hardware/software/user error as all of my machines are affected. I just want to be sure I'm not being the usual end user here and the actual problem is between the keyboard and the back of my chair ;-)

If no improvement in 12 hours I'll post back.

Thanks,

D
Team USA

PS. On my main rig I noticed the last remaining task has been running for 47 hours and reports 83 hours to completion and the time to completion is going UP.

Name: rs_stg0_lrlx_t315_run1_SAVE_ALL_OUT_19359_6103_1

Have two other tasks to complete. Then will reset to purge what appears to be a bad work unit.

This is just FYI.

PSS. In the time it took me to post this, 2 of the 3 machines poreviously reported have downloaded fresh tasks. I'm still waiting for the main machine to update. It appears things are just a little slow - not stopped.






©2023 University of Washington
https://www.bakerlab.org