AWS Thinkbox Discussion Forums

Igore Bad Slave Error Limit?

We have a mixed render farm, most machines have everything installed, but a few don’t have max 2014…
So if a user submits a max 2014 job, those machines quickly rack up enough errors to stop the job :confused:

What is the easiest way (that do not require user intervention) to avoid this from happening?

We use groups to choose between fast and slow slaves and workstations, so groups are out…

The setting “Ignore Bad Slave Error Limit” sounds like it could be a solution, but I can’t find anywhere to make it the default option for all jobs?

Any suggestions are welcome :slight_smile:
ps we use 7.2

I guess your submitting 3dsMax jobs to Deadline via SMTD?

How about creating a “limit group” and with this applied, only those machines which are present (whitelisted slaves) or alternatively (blacklisted slaves) will be able to dequeue (pick up) this job in your farm?
docs.thinkboxsoftware.com/produc … -ref-label

The limits would then be available in SMTD here:
docs.thinkboxsoftware.com/produc … limits-tab

docs.thinkboxsoftware.com/produc … mit#limits

And, if you wanted to take it to the next level, you could create an OnJobSubmitted callback that recognizes Max 2014 jobs and automatically assigns to the Job the Limit Group that Mike was suggesting. Then users would not have to explicitly assign the Limit Group at submission time.

thanks guys … this sounds like a good plan, will try and see if we can make it work :slight_smile:

Damm it Coulter! More smarts than me here. :laughing:
EDIT: although I’ll raise you and provide an example set limits event plugin all ready to go. Just need to add a line or 2, to filter against the job type and Max 2014.
github.com/ThinkboxSoftware/Dea … etJobLimit

I love being the dumbest guy in the room :slight_smile: you guys rock!

Nicely done, sir! :slight_smile:

Privacy | Site terms | Cookie preferences