Re: [PERFORM] insert waits for delete with trigger

Litao Wu <[EMAIL PROTECTED]> writes: > Did I miss something? Your join omits all transaction locks. regards, tom lane ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]

Re: [PERFORM] insert waits for delete with trigger

Hi Tom, No row has granted='f'. The result shown in the original email is from: select c.relname, l.pid, l.mode, l.granted, current_query from pg_locks l, pg_class c, pg_stat_activity a where relation is not null AND l.relation = c.oid AND l.pid = a.procpid AND l.mode != 'AccessShareLock' o

Re: [PERFORM] insert waits for delete with trigger

Litao Wu <[EMAIL PROTECTED]> writes: > Here is info from pg_lock: All those locks are already granted, so they are not much help in understanding what PID 18951 is waiting for. What row does it have with granted = 'f' ? regards, tom lane ---(end o

[PERFORM] insert waits for delete with trigger

Hi all, We have table q_20040805 and a delete trigger on it. The delete trigger is: update table q_summary set count=count-1... When we delete from q_20040805, we also insert into related info q_process within the same transaction. There is a PK on q_process, but no trigger on it. No FK on eith

Re: [PERFORM] Help specifying new machine

Rory, Rory Campbell-Lange wrote: I'm thinking of upgrading an existing dedicated server and colocating my own server. The server is used for prototype systems running Postgresql, php and apache. The largest database is presently under 10GB and I haven't had any major performance problems. We expect

[PERFORM] Help specifying new machine

I'm thinking of upgrading an existing dedicated server and colocating my own server. The server is used for prototype systems running Postgresql, php and apache. The largest database is presently under 10GB and I haven't had any major performance problems. We expect to have to host larger data set

Re: [PERFORM] tablespaces and ramdisks

"Merlin Moncure" <[EMAIL PROTECTED]> writes: > If you use tablespaces to put a high-update, non-critical table on a > ramdisk, will updates to that table will still cause the WAL files to > sync? Sure. Postgres has no way of knowing that there's anything special about such a tablespace. > I'm lo

[PERFORM] tablespaces and ramdisks

If you use tablespaces to put a high-update, non-critical table on a ramdisk, will updates to that table will still cause the WAL files to sync? I'm looking for a way to turn off syncing completely for a table. Temporary tables do this, but they can only be accessed from a single backend. Merlin

Re: [PERFORM] The black art of postgresql.conf tweaking

On Monday 09 Aug 2004 7:58 pm, Paul Serby wrote: > I've not maxed out the connections since making the changes, but I'm > still not convinced everything is running as well as it could be. I've > got some big result sets that need sorting and I'm sure I could spare a > bit more sort memory. You cou

Re: [PERFORM] The black art of postgresql.conf tweaking

-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Thanks to everyone for there help. I've changed my postgres settings to the following max_connections = 500 shared_buffers = 1 sort_mem = 2000 effective_cache_size = 5000 The 'effective_cache_size' is just a guess, but some references suggest it so