Re: Why we really need timelines *now* in PITR

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Simon Riggs <simon(at)2ndquadrant(dot)com>
Cc: Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us>, pgsql-hackers(at)postgreSQL(dot)org
Subject: Re: Why we really need timelines *now* in PITR
Date: 2004-07-19 21:58:11
Message-ID: 399.1090274291@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Simon Riggs <simon(at)2ndquadrant(dot)com> writes:
> The crucial time is when re-running recoveries repeatedly and if we
> write the manual with sufficient red ink then we'll avoid this. But
> heck, not having your history file is only as bad as not having added
> timelines in the first place. Not great, just more care required.

Yeah, you only really need them when you are hip-deep in repeated
recovery retries.

If you haven't gotten to my later proposal yet, the history files will
be plain text and it'd be at least theoretically possible for someone to
reconstruct one by hand if needed. All you need to have is the sequence
of parent timeline IDs, which you could reconstruct in most cases by
looking at the archived WAL files.

regards, tom lane

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Simon Riggs 2004-07-19 21:58:16 Re: Why we really need timelines *now* in PITR
Previous Message Simon Riggs 2004-07-19 21:50:05 Re: Why we really need timelines *now* in PITR