-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On 05/25/2011 02:47 PM, Andrew Sullivan wrote:
> On Wed, May 25, 2011 at 01:37:47PM +0100, Simon Riggs wrote:
>>
>> That's the way SQLServer and Oracle work, but not PostgreSQL. We can
>> clear down WAL files even during a long running transaction.
>>
>> For us, "unneeded" means prior to the second-to-last checkpoint record.
>
> Well, they're obviously not getting cleared down, so they must be
> needed ......
I wonder if full_page_writes has something to do with this.
- From the documentation:
".... If full_page_writes is set (as is the default), there is another
factor to consider. To ensure data page consistency, the first
modification of a data page after each checkpoint results in logging the
entire page content ...."
It looks like we get this 'problem' when creating a GIN index on a
tsvector column on a 4217 MB table (This takes longer than
checkpoint_timeout)
- --
Rafael Martinez Guerrero
Center for Information Technology
University of Oslo, Norway
PGP Public Key: http://folk.uio.no/rafael/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.14 (GNU/Linux)
iEYEARECAAYFAk3dA14ACgkQBhuKQurGihTqtgCdH+4trtuHqqqq+x0CBkYj5Vth
ZFMAninQQqtE4+ZBOA/Gff+v1Zm8xo73
=1Tbn
-----END PGP SIGNATURE-----