On 11-Jan-07, at 4:47 PM, James Neff wrote:
> Greetings,
>
> I have 3 clients trying to do a select on the same table (different
> rows). pgAdmin III Server Status shows the queries and has as mode
> "Exclusive Lock" for all three.
Are you doing select for update ? If so why so many rows ?
>
> After a long while (10 minutes maybe?) things eventually get
> unstuck and everything is alright.
>
> These clients are selecting 2000 rows from the database. My
> suspicion is the database servers IO is a bottleneck (running SATA
> drives, not SCSI). Would fetching less number of rows at a time be
> better? How can I find out what the best number of rows to fetch
> at a time are?
>
> Is this a deadlock situation?
no
>
> How would I go about diagnosing this situation, and how do I fix it?
Have you tuned the server ?
>
> I'm just using the postgresql-8.2-504.jdbc2.jar on all three clients.
>
> I'm also researching and trying to implement the pooling suggested
> before (trying DBCP first). If that will fix this problem then
> just disregard this.
>
> Thanks in advance,
> James
>
> ---------------------------(end of
> broadcast)---------------------------
> TIP 5: don't forget to increase your free space map settings
>