AWS Thinkbox Discussion Forums

beta 13, monitor slave failed tasks column

This column seems to be showing an incorrect value.

We just had a bug ripple through our whole farm causign 5-10 errors per slave, and most of them just show 0 in the monitor. When i right click and get ‘slave reports’, it shows all the error reports, but the failed tasks column value is still at 0.

Thanks! We’ve logged this as a bug.

Hmm, I can’t seem to reproduce this one…

I launched a slave and made it render a job that continuously reported errors, and the slave’s Failed Task count matched the job’s error count when all was said and done.

hmm weird… its pretty prevalent here… attahced is a screenshot, the popup slave reports is for the selected slave. The column of all the 0s is the failed task column.
failed_tasks.jpg

Note that the Failed Tasks column is only for that slave’s current session. When the slave is restarted, that value gets reset back to 0. It’s completely disconnected from the slave reports (which is the same behavior from previous versions of Deadline).

Since you restarted all of your slaves when you upgraded to beta 13, that would explain why they were all reset.

Cheers,

  • Ryan

I see…

Is there a column to show the all-time errors? That would help identify problem slaves.

There isn’t. Also, the slaves only keep track of their last 256 error reports, so the number would be capped anyways. The Failed Task column should still give a good indication of problematic slaves though.

Privacy | Site terms | Cookie preferences