[PATCH v3 07/11] replication: rebootstrap instance on startup if it fell behind

[Konstantin Osipov]

* Vladimir Davydov <vdavydov.dev at gmail.com> [18/07/19 14:07]:
> 
> What about this?
> 
> : 2018-07-19 12:53:45.195 [19455] main/101/replica I> can't follow [::1]:47383: required {1: 8} available {1: 12}
> : 2018-07-19 12:53:45.195 [19455] main/101/replica C> replica is too old, initiating rebootstrap
> : 2018-07-19 12:53:45.195 [19455] main/101/replica I> bootstrapping replica from [::1]:47383

It would be better if we use uuid or server id as identifiers
(can't follow <guid> or rebootstrapping from <guid>), providing
URIs only as additinal information.

> : 2018-07-19 12:53:46.546 [19493] main/101/replica I> can't follow [::1]:47383: required {1: 17, 2: 1} available {1: 20}
> : 2018-07-19 12:53:46.546 [19493] main/101/replica I> can't rebootstrap from [::1]:47383: replica has local rows: local {1: 17, 2: 1} remote {1: 23}
> : 2018-07-19 12:53:46.546 [19493] main/101/replica I> recovery start
> 
> The diff is below. Note, apart from logging I also changed the logic
> of the rebootstrap trigger a bit: now we proceed to rebootstrap only
> if none of the configured masters has rows needed by the instance.

Ack.

> 
> I haven't pushed the patch to the branch yet. If you're okay with it,
> I'll squash it into the original patch, rebase the whole series, and
> then update the remote branch.

OK, the patch itself is OK.



-- 
Konstantin Osipov, Moscow, Russia, +7 903 626 22 32
http://tarantool.io - www.twitter.com/kostja_osipov