Tom, Jan,
> MVCC does not mean "no locks" ... particularly not when UPDATEs are
> involved. You'll need to be more specific about what your function
> is
> doing, but my first thought would be to look for the possibility of
> conflicting updates of the same row.
Found the problem after 3 hours and a phone call to PostgreSQL Inc.
Turns out that I had one too-broad UPDATE in a 250-line function ...
feh!
-Josh
______AGLIO DATABASE SOLUTIONS___________________________ Josh Berkus Complete
informationtechnology josh@agliodbs.com and data management solutions (415) 565-7293 for law firms, small
businesses fax 621-2533 and non-profit organizations. San Francisco