Re: REINDEX on large DB vs. DROP INDEX/CREATE INDEX

From: Wes <wespvp(at)msg(dot)bt(dot)com>
To: Vivek Khera <vivek(at)khera(dot)org>, pgsql general <pgsql-general(at)postgresql(dot)org>
Subject: Re: REINDEX on large DB vs. DROP INDEX/CREATE INDEX
Date: 2008-02-05 00:37:54
Message-ID: C3CD0C02.710BA%wespvp@msg.bt.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general pgsql-hackers

On 2/4/08 9:53 AM, "Vivek Khera" <vivek(at)khera(dot)org> wrote:

> what you need to do is compare the relpages from the pg_class table
> for that index before and after.
>
> if you didn't get much disk space back, make sure you have no long
> running transactions that may have kept some older files open.

I can check that next time, but I only reindex about once a year. There
definitely should be no outstanding transactions.

The reason for the huge change in the vacuum time is that the indexes are
scanned in index order instead of disk order. I understand that is fixed in
8.2 or 8.3 (don't recall which I saw it in), but have never gotten
confirmation from anyone on that.

Wes

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Gurjeet Singh 2008-02-05 00:39:11 Re: Reload only specific databases from pg_dumpall
Previous Message Tom Lane 2008-02-05 00:12:17 Re: encoding does not match server's locale

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2008-02-05 00:55:56 Re: REINDEX on large DB vs. DROP INDEX/CREATE INDEX
Previous Message Tom Lane 2008-02-05 00:21:38 Re: TODO item:Allow to_date() and to_timestamp() accept localized month names