From: Kirill Yukhin <kyukhin@tarantool.org>
To: sergepetrenko <sergepetrenko@tarantool.org>
Cc: v.shpilevoy@tarantool.org, tarantool-patches@dev.tarantool.org
Subject: Re: [Tarantool-patches] [PATCH v2 0/5] introduce anonymous replicas
Date: Mon, 30 Dec 2019 08:12:23 +0300 [thread overview]
Message-ID: <20191230051223.3ovqiu42qzhlr3vw@tarantool.org> (raw)
In-Reply-To: <cover.1577277455.git.sergepetrenko@tarantool.org>
Hello,
On 25 дек 15:46, sergepetrenko wrote:
> https://github.com/tarantool/tarantool/tree/sp/gh-3186-anon-replica
> https://github.com/tarantool/tarantool/issues/3186
>
> The first patch alters the comment regarding join protocol
> to be in sync with latest changes: join isn't done off a snapshot,
> but from a read-view rather, created at the moment the request is received.
>
> The second patch removes unnecessary decoding of replicaset uuid on
> master side, since master just ignores it now.
>
> The third patch is a preparation for anonymous replica. We need to split join
> protocol into two stages: first, fetch a snapshot. This is done by both anonymous
> and regular replicas. The second stage is registration, i. e. the addition of
> replica to _cluster table. It is done right away when joining a regular replica,
> or it can be done by a request from an anonymous replica to transition it to
> normal one.
> The patch itself just splits applier join into appropriate stages.
>
> The fourth patch modifies vclock logic to ignore 0 component in comparisons.
> This is also used by anonymous replicas, which will be the first users of the
> 0-th vclock component. This component will be used to keep track of anonymous
> replica local changes, which are not replicated. These local changes will vary
> from instance to instance, and the anonymous instances will all use 0th component
> to track them, so no need to compare this component.
>
> The fifth patch does the main job of adding anonymous replicas.
> New requests, FETCH_SNAPSHOT and REGISTER are introduced. Anonymous replicas
> are allowed to have a 0 id. They aren't registered as gc consumers and aren't
> present in _cluster table.
>
> Changes in v2:
> - review fixes as per reviews from
> Georgy Kirichenko and Vlad Shpilevoy.
I've checked your patchset into master.
--
Regards, Kirill Yukhin
prev parent reply other threads:[~2019-12-30 5:12 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-25 12:46 sergepetrenko
2019-12-25 12:46 ` [Tarantool-patches] [PATCH v2 1/5] box: update comment describing join protocol sergepetrenko
2019-12-25 12:46 ` [Tarantool-patches] [PATCH v2 2/5] replication: do not decode replicaset uuid when processing a subscribe sergepetrenko
2019-12-25 12:47 ` [Tarantool-patches] [PATCH v2 3/5] applier: split join processing into two stages sergepetrenko
2019-12-25 12:47 ` [Tarantool-patches] [PATCH v2 4/5] vclock: ignore 0th component in comparisons sergepetrenko
2019-12-25 16:00 ` Vladislav Shpilevoy
2019-12-27 18:42 ` Vladislav Shpilevoy
2019-12-28 11:21 ` Sergey Petrenko
2019-12-25 12:47 ` [Tarantool-patches] [PATCH v2 5/5] replication: introduce anonymous replica sergepetrenko
2019-12-25 18:22 ` Vladislav Shpilevoy
2019-12-27 15:27 ` Sergey Petrenko
2019-12-27 18:42 ` Vladislav Shpilevoy
2019-12-28 11:48 ` Sergey Petrenko
2019-12-28 12:15 ` Vladislav Shpilevoy
2019-12-30 5:12 ` Kirill Yukhin [this message]
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=20191230051223.3ovqiu42qzhlr3vw@tarantool.org \
--to=kyukhin@tarantool.org \
--cc=sergepetrenko@tarantool.org \
--cc=tarantool-patches@dev.tarantool.org \
--cc=v.shpilevoy@tarantool.org \
--subject='Re: [Tarantool-patches] [PATCH v2 0/5] introduce anonymous replicas' \
/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