Tarantool development patches archive
 help / color / mirror / Atom feed
From: Vladislav Shpilevoy <v.shpilevoy@tarantool.org>
To: Sergey Ostanevich <sergos@tarantool.org>
Cc: tarantool-patches@dev.tarantool.org
Subject: Re: [Tarantool-patches] [PATCH v2 04/10] replication: track registered replica count
Date: Tue, 8 Sep 2020 00:54:51 +0200	[thread overview]
Message-ID: <bc9b194c-f9a8-3f17-0890-b5e3968a458d@tarantool.org> (raw)
In-Reply-To: <20200907154544.GA46@tarantool.org>

Hi! Thanks for the review!

>> diff --git a/src/box/replication.cc b/src/box/replication.cc
>> index ef0e2411d..20f16206a 100644
>> --- a/src/box/replication.cc
>> +++ b/src/box/replication.cc
>> @@ -247,6 +247,7 @@ replica_set_id(struct replica *replica, uint32_t replica_id)
>>  						   tt_uuid_str(&replica->uuid));
>>  	}
>>  	replicaset.replica_by_id[replica_id] = replica;
>> +	++replicaset.size;
> 
> There's another use of the replica_set_id() inside the  register_replica()
> of alter.cc. Apparently, it's just for re-assign the ID without adding of
> a new node. I would propose to move this counter into replicaset_add()

A replica can become anon, and can be deleted. So I can't manage the counter
only in replicaset_add. It should decrease somewhere.

Talking of alter.cc, instance ID is a primary key of _cluster. So it can't be
changed for an existing replica. It needs to be deleted and added with a new
ID. It means, instance ID is essentially a read-only field.

Also this is validated in replica_set_id() - it checks the replica wasn't
registered before in an assertion - and in replica_clear_id() - it checks the
replica is registered. There are no cases, when an ID is updated for an existing
replica.

  reply	other threads:[~2020-09-07 22:54 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-09-03 22:51 [Tarantool-patches] [PATCH v2 00/10] dRaft Vladislav Shpilevoy
2020-09-03 22:51 ` [Tarantool-patches] [PATCH v2 01/10] applier: store instance_id in struct applier Vladislav Shpilevoy
2020-09-04  8:13   ` Serge Petrenko
2020-09-07 22:54     ` Vladislav Shpilevoy
2020-09-03 22:51 ` [Tarantool-patches] [PATCH v2 10/10] raft: introduce box.info.raft Vladislav Shpilevoy
2020-09-04 11:38   ` Serge Petrenko
2020-09-03 22:51 ` [Tarantool-patches] [PATCH v2 02/10] box: introduce summary RO flag Vladislav Shpilevoy
2020-09-04  8:17   ` Serge Petrenko
2020-09-03 22:51 ` [Tarantool-patches] [PATCH v2 03/10] wal: don't touch box.cfg.wal_dir more than once Vladislav Shpilevoy
2020-09-04  8:20   ` Serge Petrenko
2020-09-03 22:51 ` [Tarantool-patches] [PATCH v2 04/10] replication: track registered replica count Vladislav Shpilevoy
2020-09-04  8:24   ` Serge Petrenko
2020-09-07 22:54     ` Vladislav Shpilevoy
2020-09-07 15:45   ` Sergey Ostanevich
2020-09-07 22:54     ` Vladislav Shpilevoy [this message]
2020-09-03 22:51 ` [Tarantool-patches] [PATCH v2 05/10] raft: introduce persistent raft state Vladislav Shpilevoy
2020-09-04  8:59   ` Serge Petrenko
2020-09-03 22:51 ` [Tarantool-patches] [PATCH v2 06/10] raft: introduce box.cfg.raft_* options Vladislav Shpilevoy
2020-09-04  9:07   ` Serge Petrenko
2020-09-07 22:55     ` Vladislav Shpilevoy
2020-09-03 22:51 ` [Tarantool-patches] [PATCH v2 07/10] raft: relay status updates to followers Vladislav Shpilevoy
2020-09-03 22:51 ` [Tarantool-patches] [PATCH v2 08/10] [tosquash] raft: pass source instance_id to raft_process_msg() Vladislav Shpilevoy
2020-09-04  9:22   ` Serge Petrenko
2020-09-03 22:51 ` [Tarantool-patches] [PATCH v2 09/10] raft: introduce state machine Vladislav Shpilevoy
2020-09-04 11:36   ` Serge Petrenko
2020-09-07 22:57     ` Vladislav Shpilevoy
2020-09-09  8:04       ` Serge Petrenko

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=bc9b194c-f9a8-3f17-0890-b5e3968a458d@tarantool.org \
    --to=v.shpilevoy@tarantool.org \
    --cc=sergos@tarantool.org \
    --cc=tarantool-patches@dev.tarantool.org \
    --subject='Re: [Tarantool-patches] [PATCH v2 04/10] replication: track registered replica count' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox