Exit after next checkpoint

AgentB
AgentB
Joined: 17 Mar 12
Posts: 915
Credit: 513211304
RAC: 0
Topic 197396

I´m not sure if this is a boinc or a E@H wish.

Shutdown in a frugal fashion, each task exits after it makes a checkpoint - and then boinc finishes when all tasks have stopped.

Jord
Joined: 26 Jan 05
Posts: 2952
Credit: 5893653
RAC: 76

Exit after next checkpoint

Not sure if you're reporting something, or asking for it. If you're asking for the client to shutdown after it's finished and reported all tasks, there's already an option for that.

Either in cc_config.xml set the option 1, or start the BOINC client (boinc.exe) with the --exit_when_idle attribute. This will run all tasks to their end, report them and exit the BOINC client.

See http://boinc.berkeley.edu/wiki/Client_configuration for usage.

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5872
Credit: 117634159534
RAC: 35196323

I imagine it's a new

I imagine it's a new 'feature' being asked for - close BOINC in a 'frugal' fashion. By selecting this, BOINC would monitor each running thread and close each one in turn after the next checkpoint was written. BOINC would close when all threads had closed.

I imagine this might be useful for those who shut off their computers overnight, particularly if the machine then stopped automatically after BOINC had closed. When you finished at the end of the day you could select this and walk away knowing that no loss of computation would occur for infrequently checkpointing tasks.

If BOINC had such a feature, accessible to boinccmd, it should be possible to do the job with a relatively simple script. Maybe the OP should ask for the feature over on the BOINC boards.

Cheers,
Gary.

AgentB
AgentB
Joined: 17 Mar 12
Posts: 915
Credit: 513211304
RAC: 0

Thanks Jord and Gary It´s

Thanks Jord and Gary

It´s about running a laptop or any device which is a bit slow, and has a limited amount of power or time. I know it won´t have enough power to complete a task. Some tasks are a bit long between checkpoints - FGRP3 checkpoints about 100 times per task.

I should have done a little searching first - it´s not that new a request, and looking at

http://boinc.bakerlab.org/rosetta/forum_thread.php?id=2962

and your posts gives me an idea or two to achieve a result, basically checking when

~boinc/slots/*/*.cpt files are up to date

or use

boinccmd --get_state

and compare these attributes

Eg.
checkpoint CPU time: 4227.980000
current CPU time: 4542.740000

looking for active tasks, waiting for these values to get close.

I´ll see if my modest scripting skills are up to the task.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.