hello,
We have our slaves up and running but after about a couple of hours of running they all crash. I believe this issue had something to do with what version of .Net framework was installed. Most of them have .net framework SP1 installed. All computers are experiencing the same problem though.
Any ideas??
A few questions:
- Which version of Deadline
- Which operating system
- Which version of .NET do you have installed (you mention SP1, but not if you have version 2.0, 3.5, etc installed)
Also, the next time a slave crashes, check the most recent slave log on the machine it crashed on. You can find the log folder by right-clicking on the Launcher in the system tray and selecting Explore Log Folder. Post a log here and we’ll take a look.
Cheers,
1.we are using deadline 4.0
-
we are getting the problem with machines on xp64 and win7 64
-
we have 3.5 for .net framework
here is a log of a slave that just crashed aswell
Thanks for the information. Do you .NET 2.0 SP1 installed on the machine? That’s the version of .NET that Deadline targets.
Also, there is nothing in that log that indicates a crash. However, I think that is the Launcher log, not the Slave log. The slave log will contain the word “slave” in the log file name (the launcher is the only application that prints out “Perfoming remote admin check” to its log).
Cheers,
ahh ok i think this might be the slave log
also, should we uninstall the 3 series .net frameworks?
Hmm, I wonder if it’s related to the purging of the trash, since that’s the last thing printed out. Can’t seem to reproduce this here, so it’s hard to say.
Maybe collect a couple other logs after crashes occur on different machines and see if “Purging Trash” is the last thing that they all print out.
The 3.x .NETs should be fine - you shouldn’t have to uninstall them (we have 1.1, 2.0, 3.0 and 3.5 all installed on our workstations with no issues).
heres another log with some different
I just checked a few other slaves and they do all seem to be crashing when they do purging trash.
That’s really strange that they would crash like that. The fact that there is not even an error message will likely make it very hard to track this problem down. Is this something that just started to happen recently?
One thing that’s worth trying is to open up \your\repository\trash in Windows Explorer and see what’s in the folder. If there is anything in there (files or folders), delete them all manually. After this, don’t do anything with the repository (ie: submit new jobs, delete existing ones, etc). Just let the slaves run with the repository in its current state and wait to see if they crash again when they go to purge the trash (which should remain empty).
If they do crash, then that would be quite odd because there is nothing in the trash to purge. If they don’t, then maybe it was something in trash that was causing problems, and now that you’ve cleared in out manually, things will return back to normal.
Cheers,
It looks like manually deleting the trash file did it. Thanks a lot for the help.