RFC: cluster management CLI

[knazarov]

This ticket is to prepare an RFC that describes our ideal cluster management CLI for Tarantool. Take a look at other tools like etcd, consul, and others, and come up with our own vision.

[knazarov]

• cli member

  • add
  • list
  • promote
  • remove
  • status
  • health
  • backup
  • restore

• cli user

  • add
  • delete
  • list
  • grant (role)
  • revoke (role)
  • passwd

• cli kv

  • get
  • set
  • del
  • export
  • import

• cli cluster

  • status (member/replicaset/replication status)
  • health (fail, warning, success)
  • backup
  • restore

• cli host

  • maintenance (set maintenance mode on this host, swith all masters from this node)

• cli vshard

  • create
  • delete
  • add (node)
  • delete (node)
  • weight
  • buckets (show bucket map)

• cli eval

  • member
  • replicaset
  • host
  • cluster

[tsafin]

I've no objections about management API, i.e. management of cluser, membership, host and users. But some parts are not very clear after 1st read. i.e.:

• cli kv
• cli eval

Why it should be part of cluster management?

[tsafin]

And I'd suggest to rename cli vshard to the straightforward cli shard because from the user prospective it's not actually important which implementation of sharding is used.