Re: PITR Functional Design v2 for 7.5 - Mailing list pgsql-hackers

From Tom Lane
Subject Re: PITR Functional Design v2 for 7.5
Date
Msg-id 16806.1078988259@sss.pgh.pa.us
Whole thread Raw
In response to Re: PITR Functional Design v2 for 7.5  ("Zeugswetter Andreas SB SD" <ZeugswetterA@spardat.at>)
List pgsql-hackers
"Zeugswetter Andreas SB SD" <ZeugswetterA@spardat.at> writes:
> What I think is possible is the following (continuous backup of WAL assumed):
> - disable VACUUM
> - issue CHECKPOINT "C1"
> - backup all files
> - reenable VACUUM

> - restore files
> - adapt pg_control (checkpoint "C1")
> - recover WAL until at least end of backup

I do not understand this fixation on "disable VACUUM".  What are you
thinking that will buy you?  AFAICS it would make no difference.

> I am not sure about clog, isn't clog logged in xlog ?

Right.  For the purposes of PITR, clog behaves the same as regular
data files.  You gotta back it up, but a filesystem copy will be
inconsistent until fixed by WAL replay.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: PANIC on start
Next
From: "Ramanujam H S Iyengar"
Date:
Subject: Re: How to get RelationName ??