Re: reindex/vacuum locking/performance?

From: "Matt Clark" <matt(at)ymogen(dot)net>
To: "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: "Andrew Sullivan" <andrew(at)libertyrms(dot)info>, <pgsql-performance(at)postgresql(dot)org>
Subject: Re: reindex/vacuum locking/performance?
Date: 2003-10-05 18:43:17
Message-ID: LFEIJBEOKGPDHCEMDGNFAEPOCFAA.matt@ymogen.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

> Mainly 'cause it doesn't know where the dead tuples are till it's
> looked.

At this point I feel very stupid...

> Also, VACUUM is the data collector for the free space map,
> and so it is also charged with finding out how much free space exists
> on every page.

Ah, now I just feel enlightened! That makes perfect sense. I think I had
been conflating free pages with free space, without understanding what the
difference was. Of course I still don't really understand, but at least I
now _know_ I don't.

Many thanks

Matt

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Josh Berkus 2003-10-05 18:57:21 Re: count(*) slow on large tables
Previous Message Tom Lane 2003-10-05 18:07:13 Re: reindex/vacuum locking/performance?