From: | Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com> |
---|---|
To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | Robert Haas <robertmhaas(at)gmail(dot)com>, Bruce Momjian <bruce(at)momjian(dot)us>, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: crash-safe visibility map, take three |
Date: | 2010-12-01 16:44:12 |
Message-ID: | 4CF67B5C.9030003@enterprisedb.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On 01.12.2010 18:40, Tom Lane wrote:
> Robert Haas<robertmhaas(at)gmail(dot)com> writes:
>> As far as I can tell, there are basically two viable solutions on the
>> table here.
>
>> 1. Every time we observe a page as all-visible, (a) set the
>> PD_ALL_VISIBLE bit on the page, without bumping the LSN; (b) set the
>> bit in the visibility map page, bumping the LSN as usual, and (c) emit
>> a WAL record indicating the relation and block number. On redo of
>> this record, set both the page-level bit and the visibility map bit.
>> The heap page may hit the disk before the WAL record, but that's OK;
>
> Um, no it isn't. Suppose the heap page gets to disk but we crash before
> the WAL record does. Now we have a persistent state where the heap page
> is marked PD_ALL_VISIBLE but the corresponding VM bit is not set. The
> VM bit will never become set, either, because operations on the heap
> page will see PD_ALL_VISIBLE and assume it already is set. This state
> of affairs might be acceptable from a correctness standpoint, but not
> from a performance standpoint.
The next vacuum will fix it. We already handle that. It's no different
from the situation where neither change makes it to the disk.
--
Heikki Linnakangas
EnterpriseDB http://www.enterprisedb.com
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2010-12-01 16:48:03 | Re: crash-safe visibility map, take three |
Previous Message | Tom Lane | 2010-12-01 16:43:39 | Re: crash-safe visibility map, take three |