Translating with DrWatson… this can take a few seconds the first time.
This is a genuine, complex translation. DrWatson protects commands, error codes, and log output while naturally translating the surrounding text. It’s translated once and saved.
----LNX_Mon_Mar_08_2004_12:10:03_V3.33--
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Hello everybody,
=
can anybody tell me why rollbacks on large transactions (table reorgs/loa=
ds) need
MUCH more time than they needed until the rollback started ?
=
For example a table load run for 25 minutes (not exclusive). Checkpoints =
occur (pages
are written to disk). Then the table load gets an error (long transaction=
). The following
rollback needed more than a 4 1/2 hours!
In the beginning the rollback would even have needed more time, but than =
the database
was restarted and the fast recovery was slightly faster (sure: it runexcl=
usive), but it was still
much slower than the load.
=
The data store is an HP/Hitachi XP1024 disk storage system with plenty of=
cache (read
and write) - therefore reading old log records and writing new ones (disk=
positioning)
shouldn't be too expensive - or is it?
=
I know the database is optimized for committed changes and not for rollba=
cks. Therefore
I wouldn't mind if the rollback takes two or three times as long - but it=
's much more.
=
Regards,
Andreas Kutsche
------------------------------------------------
SPAR Oesterreichische Warenhandels-AG
Hauptzentrale
Europastra=DFe 3
A-5015 Salzburg
=
Telefon : +43 662 4470 24423
E-Mail : Andreas.KUTSCHE@spar.at
Internet: http://www.spar.at
------------------------------------------------
=
----LNX_Mon_Mar_08_2004_12:10:03_V3.33----
We use strictly necessary cookies to make this site work. With your
consent we’d also use optional cookies for analytics and marketing. You can accept all,
reject all, or choose. Read our Cookie Policy.