Update operation cannot be interrupted - 'stop' action required

[Aleksey Kasatkin]

Cluster update/rollback operations cannot be stopped now. 'Stop' action is required for update/rollback operations as it's now available for deployment operation.

[Evgeniy L (rustyrobot)]

I don't think that we should implement such functional for update and rollback. (at least now)
1. new functional will add a lot of new work for our QA team
2. also from user point of view I cannot imagine case when it could be useful in fact his env will be half broken, and for patching user doesn't make any complicated configuration where he can make mistake.

[Ihor Kalnytskyi (ikalnytskyi)]

Well, what if something goes wrong and the process will never stop?
I think we should provide possibility to stop process..

[Aleksey Kasatkin (alekseyk-ru)]

1. Yes, it requires QA participation.
2. It can be useful when update/deployment time is long and user realizes that 'update' was not the right thing to do (e.g. cluster should have been redeployed with new parameters instead of being updated).

[Dmitry Pyzhov (dpyzhov)]

Stopped update/rollback will leave cluster in unpredictable state. So I'm against this feature. You can start discussion in development mailing list.