[pgpool-general: 3096] Re: prevent_data_loss

Tatsuo Ishii ishii at postgresql.org
Mon Aug 11 20:29:29 JST 2014

> This works very fine, if some of the slaves going down the data got
> immediately got restored when the slave comes back online.
> For my goal it is absolutely necessary that no data loss occurs
> whatever happens.
> I found out that data loss happens when the master goes down(crashes)
> and one of the other nodes take over.
> My question is:
> Is there a setup which prevent data loss when the master crashes?

Options are:

1) Use sync replication mode of PostgreSQL. However, you will go into
   trouble if sync standby goes down because the primary will wait
   until the sync standby comes up again. See PostgreSQL manual for
   more details.

2) Protect the primary server by using HA solution such as
   Pacemaker+DRBD. The down side of this configuration is, there will
   be DRBD replicaton overhead.

3) Use pgpool-II's native replication mode.

Best regards,
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php

More information about the pgpool-general mailing list