| From: | Scott Marlowe <scott(dot)marlowe(at)gmail(dot)com> |
|---|---|
| To: | Samuel Stearns <SStearns(at)internode(dot)com(dot)au> |
| Cc: | Jonathan Hoover <jhoover(at)yahoo-inc(dot)com>, "pgsql-admin(at)postgresql(dot)org" <pgsql-admin(at)postgresql(dot)org>, Kenneth Marshall <ktm(at)rice(dot)edu> |
| Subject: | Re: Disk Performance Problem on Large DB |
| Date: | 2010-11-05 03:41:58 |
| Message-ID: | AANLkTimVtt+XP9+-bLtLzg2WDz+BZrDjXMrpA7GTn3k+@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-admin |
On Thu, Nov 4, 2010 at 9:33 PM, Samuel Stearns
<SStearns(at)internode(dot)com(dot)au> wrote:
> TRUNCATE doesn't generate any rollback data, which makes it lightning fast. It just de-allocates the data pages used by the table.
Also truncate, like nearly everything in pgsql, can be rolled back. I
still remember showing my oracle co-dbas in my last job freak out when
I showed them things like
begin;
update table yada...
truncate table;
rollback;
or
begin;
drop index xyz;
explain select ...
rollback;
transactable everything is pretty cool. (note database and tablespace
craete / drop are the only things that aren't transactable, which
makes some sense.)
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Samuel Stearns | 2010-11-05 03:44:57 | Re: Disk Performance Problem on Large DB |
| Previous Message | Samuel Stearns | 2010-11-05 03:40:33 | Re: Disk Performance Problem on Large DB |