Job Using HOSTG Aborts and Needs to Restart on a Different Host in the Group

Discussion created by BenCroffut603954 on Apr 22, 2016
Latest reply on Apr 26, 2016 by Darren_Sniezak_40
 I've seen several discussions on here about dynamically assigning Agents in an Agent Groups to jobs at runtime. However, i do not believe this resolves the following issue:

A job is assigned to an Agent Group that contains 2+ agents. The job starts on Agent_1. For whatever reason Agent_1 goes down. The job will eventually abort (probably ended_lost).

How can we restart this job so that it gets assigned to Agent_2 or Agent_3. By default, this job when restarted will go straight to Waiting_for_host status for Agent_1.

We are moving to more of a commodity hardware infrastructure and when a host goes down, a different host should be able to pick up the work (shared filers, syncing data between hosts, etc...). So when an agent goes down, which is becoming more often with these new hosts, we'd like to reassign all these jobs to a separate host in the group.

I don't believe Automic has a solution out of the box for this but if you are aware of any I'd appreciate your input.