From: | Ed Loehr <ELOEHR(at)austin(dot)rr(dot)com> |
---|---|
To: | Mike Mascari <mascarm(at)mascari(dot)com> |
Cc: | Barnes <aardvark(at)ibm(dot)net>, pgsql-general(at)postgresql(dot)org |
Subject: | Re: [GENERAL] Re: Is PostgreSQL ready for mission criticalapplications? |
Date: | 1999-12-29 20:28:22 |
Message-ID: | 386A6EE6.87DD3C82@austin.rr.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Thanks, Mike! This is the most lucid, concise explanation of so many
postgresql "gotchas" I've seen yet.
Mike Mascari wrote:
> 2. Using views created with large queries - Views use the
> rewrite system and rules to rewrite a query against it to
> properly fetch data from the underlying tables. Because
> there is currently a limit on the size of a single database
> record (8192 bytes), the queries associated with views can
> only be so big. ...
One additional anomaly as of 6.5.2 regarding backup and recovery...
If one simply compares the before/after output of load/dump scripts, it can at
first appear that pg_dump will occasionally convert a view built on non-empty
tables into a table itself with zero records. This happens during the
following backup test sequence for me:
% pg_dump -d mydb > db.out
% destroydb mydb
% createdb mydb
% psql -d mydb < db.out
% pg_dump -d mydb > db2.out
% diff db.out db2.out
This is because a view _is_ actually implemented as a table combined with a
redirecting rule, and thus not a problem. See the following for details.
http://www.deja.com/getdoc.xp?AN=559228857
Cheers,
Ed Loehr
From | Date | Subject | |
---|---|---|---|
Next Message | Peter Landis | 1999-12-29 22:02:11 | |
Previous Message | Bruce Momjian | 1999-12-29 19:45:13 | Re: [GENERAL] Re: Is PostgreSQL ready for mission criticalapplications? |