| From: | "Scott Marlowe" <scott(dot)marlowe(at)gmail(dot)com> |
|---|---|
| To: | "Mark Roberts" <mailing_lists(at)pandapocket(dot)com> |
| Cc: | "Peter Billen" <peter(at)clueless(dot)be>, pgsql-general(at)postgresql(dot)org |
| Subject: | Re: SERIAL datatype |
| Date: | 2008-08-25 17:02:37 |
| Message-ID: | dcc563d10808251002s37b9cd08m491ec969fb5cfdc2@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
On Mon, Aug 25, 2008 at 10:23 AM, Mark Roberts
<mailing_lists(at)pandapocket(dot)com> wrote:
>
> On Thu, 2008-08-21 at 13:53 -0600, Scott Marlowe wrote:
>> Regular SERIAL type is limited to a 32 bit int. BIGSERIAL uses a 64
>> bit int.
>
> I think one of the things that would be offsetting is the size
> difference between the two types (32 vs 64 bits, 5 foreign keys, and a
> billion rows or so makes for alot of pain).
Well, of course a 64 bit int is gonna be bigger than a 32 bit, but
with alignment issues and on 64 bit hardware, I'm guessing the
difference isn't exactly twice as slow / twice as much storage. And
it's way faster than a GUID which was what I think started this
thread.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Joshua Drake | 2008-08-25 17:04:13 | Re: Dump/restore with bad data and large objects |
| Previous Message | John T. Dow | 2008-08-25 16:58:00 | Re: Dump/restore with bad data and large objects |