[rt-users] postgres slowdowns? (pg 7.3.4, rt 3.0.4)

Justin Hawkins justin at internode.com.au
Sun Oct 12 23:13:19 EDT 2003


On Thu, Sep 25, 2003 at 10:20:16AM -0400, Vivek Khera wrote:
> >>>>> "JH" == Justin Hawkins <justin at internode.com.au> writes:
> 
> JH> Postgres doesn't actually seem to think this is a particularly
> JH> stressful query:
> 
> JH> Limit  (cost=531.38..531.45 rows=1 width=212)
> 
> JH> But it is - execution time is: 47218.74 ms !!!
> 
>  [...]
> 
> JH> If I alter the above query to say (main.Status = 'open') instead of
> JH> 'new' OR 'open', the query drops to about 125ms :-)
> 
> Since the quantity of data has changed drastically, have you
> re-computed the analyzer stats by running "vacuum analyze" on your
> database?  What is the execution plan from postgres for the above
> queries (both before and after "analyze")?

Sorry for the delay - been on leave :-)

Yes - I always analyse when in doubt. It makes no appreciable difference to
either the query cost, or actual runtime.

	- Justin



More information about the rt-users mailing list