Long running work unit

[trindol]

6077994
6083144

I am concerned with these two. When they were at 300 hours, the host lost power. The units restarted automatically when power was restored. Now they are at 2400 hours. Both have been at 100% for weeks. Usually this host is much faster even with XXL work units. Do you think these units will be ok despite the crash? Thanks.

[Michael H.W. Weber]

The crash just caused a regular restart.
By the way: We have a VM-based checkpointing system now.

Michael.

[trindol]

Those two work units need an extension please.

[yoyo]

extended.

[rebel9]

I also have this one:

http://www.rnaworld.de/rnaworld/result. ... d=14920814

which is overdue. It's been running for 9.5 days and been running at high priority for some time. It has been pegged at the 98.765 limit for ages and the time left has been at around 7 hrs for as long as I can remember.

What should I do with it?

Thanks.

[yoyo]

This task isn't overdue. According to the server the deadline is 9 Dec 2013, 20:00:49 UTC. So let it run.

[rebel9]

Hi,

All my BOINC clients are showing that the deadline was 9/11/2013. I hadn't noticed that the site showed a different date and I'm assuming that's the correct one as it's the only active one and the only VBox WU. If so, there's a problem with your WUs. I'm happy to let it run for now, though.

Thanks.

[ChristianB]

Hi,

All my BOINC clients are showing that the deadline was 9/11/2013. I hadn't noticed that the site showed a different date and I'm assuming that's the correct one as it's the only active one and the only VBox WU. If so, there's a problem with your WUs. I'm happy to let it run for now, though.

Thanks.

There is no problem per se with our WUs. Just the one you have is from a batch where automatic deadline extension and the progress indicator aren't working. The deadline was extended on the server side. The newer tasks don't have this problem so you may cancel this WU and get a new one in the future (that has automatic deadline extension and a working progress bar) if you want.

The deadline on the Client won't change because BOINC doesn't support updating the deadline. Also we are still trying to find bugs in the VM and create new WUs only if needed for the moment. All our tasks are known to exceed the initial deadline but as long as you let it finish you will successfully contribute to RNA World.

[rebel9]

OK, that's good to know, thanks.

[candido]

I have this task that has been runing for 280 hours and is showing 30% compelete.

wu 6330294
cms_GA-p[b-Lin64f-2]_1_Tolumonas-auensis-DSM-9187_CP001616.cir.EMBL_RF00028_Intron_gpI_1358679723_2015

The deadline is 01/December/2013.
Could you please extende the deadline?
Thanks

[Jacob Klein]

Could you please carefully extend the deadline for task 14921181 (from 12/12/2013 to maybe 1/1/2014)?
http://www.rnaworld.de/rnaworld/result. ... d=14921181

Status:
It is a VM task, I am 740 hours in, it is running high-priority nearly 24/7, and because it is one of the old v1.03 tasks, I believe automatic deadline extension is not working.
I am carefully babysitting it - monitoring stderr.txt, and when I see it grow to a couple MB, I close BOINC, wait for the processes to stop, delete the content of stderr.txt, and restart BOINC.
I'm hopeful it will complete successfully, eventually; estimated runtime on reference system is 1447 hours.

Thanks,
Jacob

[ChristianB]

extended