Also,
- who will add "hot_standby = on" in the new Standby and remove it from new
Master?
- who will add "host replication..." entry in the new Standby's pg_hba.conf?
I didn't find all these settings in basebackup. Though, we handle some of them.
Please help.
________________________________
From: Sandeep Thakkar <[email protected]>
To: Sandeep Thakkar <[email protected]>; Tatsuo Ishii <[email protected]>
Cc: "[email protected]" <[email protected]>
Sent: Thursday, September 22, 2011 8:10 PM
Subject: Re: [Pgpool-general] Is pgpool_walrecrunning needed in Master/Slave
mode?
I know there are lot of posts on online recovery and I have gone through them,
but still I've some questions which are unanswered. They are:
In streaming replication mode, I have one Master (5432), and one Standby(5433),
now once the failover happens, I observe the following:
- cluster directory of new Master (5433) contains recovery.done
- postgresql.conf of new Master (5433) still contains "hot_standby = on".
Is it fine?
Now, I want to bring up the original Master (5432) as a new Standby. Hence, I
create basebackup and pgpool_remote_start scripts in the cluster directory of
new Master(5433). Before I run pcp_recovery_node, how about the following
things?
- existing cluster directory of old Master (5432). Who will remove it and when?
- recovery.done in cluster of new Master (5433). This needs to be removed Or?
If yes, who will remove it?
- where will recovery.conf will reside now? in cluster directory of 5432 or
5433?
Thanks.
________________________________
From: Sandeep Thakkar <[email protected]>
To: Tatsuo Ishii <[email protected]>
Cc: "[email protected]" <[email protected]>
Sent: Thursday, September 22, 2011 7:02 PM
Subject: Re: [Pgpool-general] Is pgpool_walrecrunning needed in Master/Slave
mode?
Thanks Guys.
Since I use pgpool-II 3.0.4, I installed this function. and when I try to
recover the original primary which was down, it recovers, but comes up as
primary. It should actually be a standby, right?
I also the following line in pgpool.log:
Replication of node:1 is behind 67109104 bytes from the primary server (node:0)
It seems, I have done something wrong. Any clues here?
Thanks.
________________________________
From: Tatsuo Ishii <[email protected]>
To: [email protected]
Cc: [email protected]
Sent: Saturday, September 17, 2011 1:40 PM
Subject: Re: [Pgpool-general] Is pgpool_walrecrunning needed in Master/Slave
mode?
> I read the 3.0.3 docs and found that pgpool_walrecrunning() should be used
> only in replication mode, and that too when somebody is using the
> failure/online recovery feature of pgPoll-II.
pgpool_walrecrunning is only usefull for streaming replication
mode in 3.0.x.
> At the same time, I found that this function has undergone some change:
> "
>
> Now "pgpool_walrecrunning()" was not used. pgpool-II used to consider the
> node that is promoted a primary node using the function. Now, pgpool-II waits
> for completing of the promotion to primary node because it did not work as we
> intended. But we still have a problem that
pgpool-II
waits while recovery_timeout, when there is no primary node(Kitagawa)
> "
>
> Can someone elaborate this please? Also, I assume that I need this function
> only when I setup pgPool-II in replication mode.
In pgpool-II 3.1 the function is not used anymore even in streaming
replication mode.
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese: http://www.sraoss.co.jp
_______________________________________________
Pgpool-general mailing list
[email protected]
http://pgfoundry.org/mailman/listinfo/pgpool-general
_______________________________________________
Pgpool-general mailing list
[email protected]
http://pgfoundry.org/mailman/listinfo/pgpool-general
