[postgis-users] ST_SPHEROID_LENGTH SLOW
mark.cave-ayland at siriusit.co.uk
Tue Oct 7 13:52:34 PDT 2008
David Vaz wrote:
> I am unable to find any significant difference in the configurations,
> and the 2 datasets are in fact equal. It might be a slight difference in
> the configuration or even in compilation of postgres, but I am no expert
> in this so I need some help. I have attached both configuration and
> output of 'SHOW ALL'. I hope someone can give light on this subject. To
> me it seems the problem is on the Postgres side as you pointed out.
If the two configurations are equal, I wouldn't expect too much
difference between the installations. The key part really is related to
the vacuuming; 8.3 vacuum is known to eliminate some vacuum starvation
cases, but this seems to be the opposite to what you are seeing.
Have you tried a pg_dump/re-initdb/pg_restore cycle on your 8.3
installation? If the problem turns out to be either corruption or
table/index bloat then this should immediately resolve it. Also just to
confirm - where are you getting your debian packages from?
Sirius Corporation - The Open Source Experts
T: +44 870 608 0063
More information about the postgis-users