you're probably loosing the network connection when you run this which is why it fails...
try scheduling the command to run w/ 'at' or something to run 1 min after the job executes.
I have tried that too... ive put in a sleep and then continue with the job, but that still fails. Whats strange is if i run that job and im logged onto the server then it works (which is no good to me) as i need it to work without logging on.
not sleep, a scheduled task. the run of the nic config command needs to happen after the bladelogic job has finished. the nic config will cause bladelogic to loose communication w/ the agent if it's run during the job and think that it's failed because of that. doing the sleep means it's still running inside the job.
hmm - in the blpackage, try using the 'out-of-band' reboot option ? whether it's just the nic going up and down or a reboot should look the same to bladelogic.
how about a sleep after?
maybe the command returns immediately, and the bldeploy finished immediately, killing all child processes.
(I got this problem running the oracle installer)
I don't think that an external command needs the network up to complete a deploy.
Did you ever get a solution to this issue?
I'm trying to do the same thing but Bladelogic thinks the job is still running yet the processes all stopped when the network disappears.
The scripts are all local at that point.
A scheduled task is a bit messy for me :-(