[ML] Allowing more job config settings to be updated on reset jobs

[droberts195]

We only allow a subset of job configuration settings to be updated after the job is created. This is because updating some of the settings would cause the job config to be out-of-sync with the results and/or state. But in the case of a job that has been reset (or never opened in the first place) there are no results or state for the config to be out-of-sync with.

Therefore, in the case of a job that has been reset (or never opened in the first place) it would be nice to allow all fields within the job config to be updated.

The remaining complication is the interaction between the job config and its associated datafeed config. Some datafeed config fields are validated against values in the associated job config, for example datafeed frequency and job bucket span. It would be easier to resolve this difficulty if the job and datafeed configs were a single object, and while this is the long-term goal, it will take a while to get there. In the short term we will need to disallow job config updates that would cause the validation of the associated datafeed to fail.

[elasticmachine]

Pinging @elastic/ml-core (Team:ML)

[EmekaMomodu]

Hello @droberts195, is this still open ? I would like to make my first contribution through this.

[EmekaMomodu]

Currently working on this.

[sharubitsi]

Hello i am willing to start my open source software development, is it possible to me work on this issue?

[droberts195]

@sharubitsi yes, if you'd like to tackle this then that would be great.