Re: Failback to old master - Mailing list pgsql-hackers

From Maeldron T.
Subject Re: Failback to old master
Date
Msg-id CAKatfS=AqZfBywvV+ZGrjhXh=1EPPawT5nj8KMPgEhVCyZBLsQ@mail.gmail.com
Whole thread Raw
In response to Re: Failback to old master  (Heikki Linnakangas <hlinnakangas@vmware.com>)
List pgsql-hackers


2014-11-13 9:05 GMT+01:00 Heikki Linnakangas <hlinnakangas@vmware.com>:

Right. You have to be careful to make sure the standby really did fully catch up with the master, though. If it happens that the replication connection is momentarily down when you shut down the master, for example, then the master won't wait for the standby. You can use pg_controlinfo to verify that, before promoting the standby.


- Heikki


Dear Heikki,

would you please tell me which line I should check to be 100% sure that everything was sent to the slave when the master was shut down?

Latest checkpoint location:           1F/B842C3D8
Prior checkpoint location:            1F/B837B9B8
Latest checkpoint's REDO location:    1F/B841A050
Latest checkpoint's REDO WAL file:    000000010000001F000000B8
Latest checkpoint's TimeLineID:       1
Latest checkpoint's PrevTimeLineID:   1
Latest checkpoint's full_page_writes: on
Latest checkpoint's NextXID:          0/15845855
Latest checkpoint's NextOID:          450146
Latest checkpoint's NextMultiXactId:  2250
Latest checkpoint's NextMultiOffset:  4803
Latest checkpoint's oldestXID:        984
Latest checkpoint's oldestXID's DB:   1
Latest checkpoint's oldestActiveXID:  15845855
Latest checkpoint's oldestMultiXid:   1
Latest checkpoint's oldestMulti's DB: 1

Is it the first line (Latest checkpoint location) or do I have to check more/else?

I plan to do this on the weekend.

Thank you.

M.

pgsql-hackers by date:

Previous
From: Tomas Vondra
Date:
Subject: Re: logical column ordering
Next
From: Tomas Vondra
Date:
Subject: Re: New CF app deployment