[pgpool-general: 5718] Re: Questions about Replication mode

Lazaro Garcia lazaro3487 at gmail.com
Fri Sep 8 22:28:02 JST 2017


1.       Is it a Master-Master Replication Type?

In replication mode, pgpool-II acts as a middleware intercepting all queries and send those queries to postgresql servers. The postgresql servers runs individually. There is no connections between then, for that reason a direct change in one node is not propagated to other nodes, and this is not a master-master replication.  Replication mode works like synchronous replication. (All nodes needs to confirm the request to pgpool, if the request fails on one node the query could by rolledback)

2.       Are there Primary and Standby or not?

All nodes are primary nodes.

3.       If a postgres server goes down and then restarts, is there an automatic way to update the data?

Is one node goes down pgpool-II detach that nodes and triggers failover command. Later is necessary an online recovery by pgpool.

4.       Is there a failover and recovery procedure?

You can configure failover_commands and recovery_commands in pgpool-II.

I think master-slave mode with postgresql replication (streaming replication) is the best choice.

Regards.

 

 

De: pgpool-general-bounces at pgpool.net [mailto:pgpool-general-bounces at pgpool.net] En nombre de Giancarlo Celli
Enviado el: viernes, 8 de septiembre de 2017 05:15 a. m.
Para: pgpool-general at pgpool.net
Asunto: [pgpool-general: 5717] Questions about Replication mode

 

Hi to All,

 

I am beginner on pgpool.

I want to understand the difference between master-slave and pgpool replication mode.

Now, I have implemented a postgresql streaming replication, so I'm using master-slave mode on pgpool-II.

However, I would like to understand the characteristics of internal replication at pgpool:

1.       Is it a Master-Master Replication Type?

2.       Are there Primary and Standby or not?

3.       If a postgres server goes down and then restarts, is there an automatic way to update the data?

4.       Is there a failover and recovery procedure?

Thanks to all.

Best regards.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.sraoss.jp/pipermail/pgpool-general/attachments/20170908/a245970c/attachment.html>


More information about the pgpool-general mailing list