Re: replication consistency checking

From: hydra <hydrapolic(at)gmail(dot)com>
To: "pgsql-admin(at)postgresql(dot)org" <pgsql-admin(at)postgresql(dot)org>
Subject: Re: replication consistency checking
Date: 2015-06-11 05:16:14
Message-ID: CAG6MAzTm4FvmENEwDtgSRCyepDNAm3wVq3b-GRG_38UZ0AjNjg@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-admin

On Tue, Jun 9, 2015 at 10:15 PM, Jan Lentfer <Jan(dot)Lentfer(at)web(dot)de> wrote:

>
>
> Am 04.06.2015 um 06:49 schrieb hydra <hydrapolic(at)gmail(dot)com>:
>
> After setting up streaming replication, is it possible to check whether
> the slave has the same data as the master?
>
> In the MySQL world there is the percona-toolkit with pt-table-checksum
> that does this job:
> https://www.percona.com/doc/percona-toolkit/2.2/pt-table-checksum.html
>
> Thank you
>
>
> I have not digged to deep, but as far as I can see from your link:
>
> *pt-table-checksum* requires statement-based replication, and it sets
> binlog_format=STATEMENT on the master
>
> And the mysql manual says
>
> If you are using InnoDB
> <https://dev.mysql.com/doc/refman/5.6/en/innodb-storage-engine.html> tables
> and the transaction isolation level is READ COMMITTED
> <https://dev.mysql.com/doc/refman/5.6/en/set-transaction.html#isolevel_read-committed>
> or READ UNCOMMITTED
> <https://dev.mysql.com/doc/refman/5.6/en/set-transaction.html#isolevel_read-uncommitted>,
> only row-based logging can be used.
>
> https://dev.mysql.com/doc/refman/5.6/en/binary-log-setting.html
>
>
> So, if I am not misinterpreting, this plays in a different league.
> Jan
>

The recommended binlog format for replication is ROW, so everything except
the checks is using ROW and only those checks have STATEMENT set.

In response to

Browse pgsql-admin by date

  From Date Subject
Next Message hydra 2015-06-11 05:24:09 Re: replication consistency checking
Previous Message hydra 2015-06-11 05:14:33 Re: replication consistency checking