Why is red a problem?
What is in the job run log of the batch job?
Thanks for the fast reply. For me Red is no Problem. But for the User who sometimes starts the Job manually.
I told them: "just look at the Time, you'll see your failed run out of all the failed Batchjob Runs."
They could just deduce it...
I'm just looking for a "nicer" Solution than the one i've implemented currently.
Inside the Batchjobrun theres the following:
Error 23.08.2012 11:30:09 No servers for batch job "DeployJobName"
I'm not sure i understand though - if there are no servers we could show a warning instead of the error. how would either one of those help the user differentiate which run is the one they are concerned with?
if you want to delete the run, you should be able right click on the run and choose 'delete' from the right most windows pane in the gui. but if the user manually runs the job, if fails because there are no targets, and you delete the run, won't that confuse the user even more than seeing a failed run?
what is the overall goal here?
Because the Job at 10:53:32 really faild and all the others "just" had no targets.
I should have specified the delete action a little more specific. I'm looking for a solution out of BLCLI.
The overall goal is: "recently added Systems get standard Software"
I wrote a Bladelogic Add Server Script. (Properties and all the ACL Stuff.) Works quiet fine, thanks for many other helpfull answers you already gave me by other posts.
But i don't want to write the Batchjob triggering logic inside the script due to maintenance reasons...
So i decided to set properties to add them into special smartgroups.
i use those smartgroups inside a scheduled batchjob to install new software.
inside the batchjob theres a property revert script. after this script has run, the Server is no longer in the Smartgroup and the "provisioning" is finished.
you may see my "beauty" problem now.
So instead of using a batch job directly, you could have a nsh job that is scheduled that checks the group to see if there are members, then if there are, run the batch job.
In this case i'd have to know about the relation between Smartgroup and Batchjob.
And as far as i know theres no possibility to get the relation out of BLCLI.
May you know how to find related batchjobs by Smartgroups, or may it's possibly to check batchjobs for targets first...?
I don't want to start all the Batchjobs, otherwise a have the same situation as before...
1 of 1 people found this helpful
there's not any blcli that will get the group target from the job. you will have to maintain some kind of association between the job and the target group if you want to have a wrapper process determine if there are targets for the smart group.
one other question - even if we get rid of all the runs w/ no targets, the user is still going to have to go through and figure out which run applies to them right?
what governs when you can run the batch jobs?
what if, when you imported the server you create an execution task for the batch job, targeted at the server(s) you imported and set a schedule for it to run at? then you could point the user at a particular ET for them to review their results?
Thanks for you're input Bill.
Yes, you're right. They'd still have to go though the Jobs and find the appropriate one. But it would be simplified if they could just ignor the one "w/ no targets"... Can you see my point?
Currently we've just reduced the triggering of the Batchjob (the Batchjobs aren't Scheduled) to run only if there are new Systems Added. So we triger the Batchjob out of the System Add NSH Script. This reduces the red bullets drasticly.