Re: Allow vacuumdb to only analyze - Mailing list pgsql-hackers

From Kevin Grittner
Subject Re: Allow vacuumdb to only analyze
Date
Msg-id 4A1D160F.EE98.0025.1@wicourts.gov
Whole thread Raw
In response to Re: Allow vacuumdb to only analyze  (Robert Haas <robertmhaas@gmail.com>)
List pgsql-hackers
Robert Haas <robertmhaas@gmail.com> wrote:
> I noticed in Bruce's talk that there are a number of post-migration
> steps which are currently partially manual.  Ideally we'd like to
> automate them all, preferably in some sort of well-thought-out
order.
> I actually suspect this is something like: analyze each database,
> reindex those indices invalidated by the upgrade, analyze each
> database again.
We have found it useful to VACUUM FREEZE ANALYZE a converted database.
The first access to any page will cause writing of hint bits, and
we'd rather deal with that before we let the users in, to avoid having
sluggish performance for them while that happens.  The FREEZE part is
to avoid a freeze of all subsequently untouched data in all tables at
some unpredictable time -- potentially in the middle of a busy
workday.  The usual argument against aggressive freezing (that
forensic information useful in recovery of a corrupted database)
doesn't carry much weight right after a conversion.
-Kevin


pgsql-hackers by date:

Previous
From: Fujii Masao
Date:
Subject: Re: Synchronous replication: Promotion of Standby to Primary
Next
From: Markus Wanner
Date:
Subject: Re: PostgreSQL Developer meeting minutes up