Give user feedback on background migration jobs

[JeffreyDevloo]

Problem description

Updating your environment will do the most necessary tasks first while the services are down but afterwards a background job is scheduled.
The only way to see if this background job has failed is parsing the logs.
The problem is that the user has no direct feedback whether his environment has been fully updated or not.

We should create a page and steps to display the full flow of the update and provide the user with more insights.

[wimpers]

Any reason why we would need a new page or can we add this information to the Update page under Administration?

[JeffreyDevloo]

It might get too cluttered in the future

[matthiasdeblock]

Why not creating a job logger like we have on CMC on VCPE? This will also make life as OPS easier to figure out on which node the job got triggered... For example, on Nuvolat we have +- 20 worker nodes...

[wimpers]

@ops can you investigate (if time permits) if there is a tool that might help. Writing this from scratch might be out of scope for a long time.

[matthiasdeblock]

The only thing I can come up with is http://flower.readthedocs.io/en/latest/. But actually I'm not sure if this is the tool that we want and if it is detailed enough as this is only for Celery.

[JeffreyDevloo]

I don't think that this is the kind of tool you want.
It's just the framework that should be storing the result of the scheduled tasks (the ones that are cronned to run) somewhere for the last x schedules
This way, we won't lose the information within the logging heap