Controlling the tasks performed by CloudStack could be interesting if something goes wrong. Even if cloudstack-management crashes, the queuing system could keep the running tasks organized and when returning to its active state again, it could resume from the previous point, also enabling the performance of rollback, etc. In case there is more than one active management, the same task could be consumed by all listeners and executed by all instead of the current model.
1 Comment
Daan
I am not sure what we need here Marco Sinhoreli . i'm sure you have a good use-case though. lets discuss furthe some day.