want to expose "force-master-takeover" by http api & orchestrator-client

[MOON-CLJ]

want to expose "force-master-takeover" by http api & orchestrator-client。only available in cli currently。

[shlomi-noach]

force-master-failover exists in API and in orchestrator-client.

API:

• force-master-failover/:host/:port
• force-master-failover/:clusterHint

orchestrator-client:

• orchestrator-client -c force-master-failover

$ orchestrator-client -c help | grep force
    force-master-failover                    Forcibly discard master and initiate a failover, even if orchestrator doesn't see a problem. This command lets orchestrator choose the replacement master

[MOON-CLJ]

@shlomi-noach i mean "force-master-takeover" not "force-master-failover"

[shlomi-noach]

Ah, sorry, I wasn't reading this right.
Yes, it's possible to add to api and to orchestrator-client, though I've been having thoughts on the usefulness of this command. I never use it, and have instead opted to use either graceful-master-takeover or force-master-failover.

force-master-takeover is for when a recovery should be taking place, but is not, and you want to force it, and you also indicate where to failover too.
In my current thinking, is something like this happens I'd rather orchestrator just failed over where it needs to, and the burden of picking a specific destination is a harsh constraint during an outage situation.

Does that makes sense?

[MOON-CLJ]

@shlomi-noach such as a situation，the old master is very busy，we cant execute GracefulMasterTakeover，but we also want to issue a destination to promote，so we can only use ForceMasterTakeover。

[shlomi-noach]

Very good. I'll add the API

[MOON-CLJ]

@shlomi-noach thx! if you don't have much time, we can do this。we already have a commit to do this，should we give a pull request？

[shlomi-noach]

Yes, please.

[MOON-CLJ]

thx