I have a question regarding "connection for xxyy established" The situation
below shows records being added to 3 tables which are heavily populated. We
never "update" any table, only read from them. Or we delete a full-day worth
of records from them.

The question is: Is this method of repeatedly establishing and
re-establishing database connections with the same 3 tables efficient? As
in, is there a better way to add data?

Apr 25 03:38:47 saw alert_mgr[33696]: Database connection for Tbl_A
established
Apr 25 03:38:51 saw alert_mgr[33698]: Database connection for Tbl_B
established
Apr 25 03:38:54 saw alert_mgr[33700]: Database connection for Tbl_A
established
Apr 25 03:38:55 saw alert_mgr[25182]: user_new_locked: added 64 entries
(66880 total)
Apr 25 03:38:57 saw alert_mgr[33702]: Database connection for Tbl_B
established
Apr 25 03:38:59 saw alert_mgr[33704]: Database connection for Tbl_A
established
Apr 25 03:39:02 saw alert_mgr[33706]: Database connection for Tbl_B
established
Apr 25 03:39:04 saw alert_mgr[33708]: Database connection for Tbl_C
established
Apr 25 03:39:05 saw alert_mgr[33710]: Database connection for Tbl_A
established
Apr 25 03:39:06 saw alert_mgr[25182]: user_new_locked: added 64 entries
(66944 total)
Apr 25 03:39:06 saw alert_mgr[33712]: Database connection for Tbl_B
established
Apr 25 03:39:08 saw alert_mgr[33714]: Database connection for Tbl_A
established
Apr 25 03:39:11 saw alert_mgr[33716]: Database connection for Tbl_B
established
Apr 25 03:39:13 saw alert_mgr[33718]: Database connection for Tbl_A
established
Apr 25 03:39:15 saw alert_mgr[25182]: user_new_locked: added 64 entries
(67008 total)
Apr 25 03:39:18 saw alert_mgr[33720]: Database connection for Tbl_B
established

--
Yudhvir Singh Sidhu
408 375 3134 cell

"Y Sidhu" <ysidhu@gmail.com> writes:
> The question is: Is this method of repeatedly establishing and
> re-establishing database connections with the same 3 tables efficient?

No. Launching a new backend process is a fairly expensive proposition;
if you're striving for performance you don't want to do it for just one
or two queries. Look into connection pooling ...

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 1: if posting/reading through Usenet, please send an appropriate
subscribe-nomail command to majordomo@postgresql.org so that your
message can get through to the mailing list cleanly

You are referring to pgpool? BTW, thanks for this insight.

Yudhvir
========

On 5/30/07, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>
> "Y Sidhu" <ysidhu@gmail.com> writes:
> > The question is: Is this method of repeatedly establishing and
> > re-establishing database connections with the same 3 tables efficient?
>
> No. Launching a new backend process is a fairly expensive proposition;
> if you're striving for performance you don't want to do it for just one
> or two queries. Look into connection pooling ...
>
> regards, tom lane
>



--
Yudhvir Singh Sidhu
408 375 3134 cell