| From: | "Henry Ortega" <juandelacruz(at)gmail(dot)com> |
|---|---|
| To: | pgsql-sql(at)postgresql(dot)org |
| Subject: | Make Trigger run after completion of ENTIRE transaction |
| Date: | 2006-09-19 19:00:52 |
| Message-ID: | 2bffcc330609191200l5e84ec23x8db6b7beb89532e5@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-sql |
This maybe more of a theoretical question, can you actually make a Trigger
run
after completion of the entire transaction?
Here's what I have:
LOG
user | startdate | enddate
enddate is getting updated by a trigger (on insert or update).
I have the following transaction:
BEGIN;
Insert into LOG(user,startdate) values('jdoe','2006-08-13');
Insert into LOG(user,startdate) values('jdoe','2006-08-14'); Insert into
LOG(user,startdate) values('jdoe','2006-08-15'); Insert into
LOG(user,startdate) values('jdoe','2006-08-16'); Insert into
LOG(user,startdate) values('jdoe','2006-08-17');
........... *another 20-30 more inserts..........*
COMMIT;
The trigger actually runs on each Insert and therefore slows down the
Insert quite a bit.
My question is, can you tell the trigger to run after Commit?
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tom Lane | 2006-09-19 19:15:09 | Re: Nested loops are killing throughput |
| Previous Message | CG | 2006-09-19 18:59:13 | Re: Nested loops are killing throughput |