Postgresql-8.3.1

I have a plpgsql function which is called nightly to update rows in a summary
table. The summary table has foreign keys that reference the users table.
When the nightly job runs, the users table gets locked such that UPDATES to
the user table are stuck waiting on the transactionid of the function.

The function does not update or select for update any rows in the users table
and removing the foreign keys in the summary table allows the UPDATES of the
users table to happen with no problem.

Interestingly, in testing this, I started a transaction, updated a bunch of
rows in the summary table with UPDATE, left the transaction open, started
a new session and successfully updated rows in the users table that are
pointed to by the updated rows in the summary table. So there doesn't seem to
be a locking problem outside of the function.

Is the locking behavior different inside a plpgsql function?

--
Jeff Frost, Owner <jeff@frostconsultingllc.com>
Frost Consulting, LLC http://www.frostconsultingllc.com/
Phone: 650-780-7908 FAX: 650-649-1954

--
Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-admin