From: | Scott Marlowe <smarlowe(at)g2switchworks(dot)com> |
---|---|
To: | Michael Fuhr <mike(at)fuhr(dot)org> |
Cc: | KÖPFERL Robert <robert(dot)koepferl(at)sonorys(dot)at>, "'pgsql-sql(at)postgresql(dot)org'" <pgsql-sql(at)postgresql(dot)org> |
Subject: | Re: Column with recycled sequence value |
Date: | 2005-01-13 21:31:54 |
Message-ID: | 1105651913.24795.102.camel@state.g2switchworks.com |
Views: | Whole Thread | Raw Message | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-sql |
On Thu, 2005-01-13 at 15:19, Michael Fuhr wrote:
> On Thu, Jan 13, 2005 at 02:48:47PM -0600, Scott Marlowe wrote:
> > On Thu, 2005-01-13 at 11:08, KÖPFERL Robert wrote:
>
> > > suppose I have a let's say heavy used table. There's a column containing
> > > UNIQUE in4
> > > values. The data type musn't exceed 32-Bit. Since however the table is heavy
> > > used 2^32 will be reached soon and then? There are far less than 4G-records
> > > saved thus these values may be reused. How can this be accomplished?
> > >
> > > I already defined a sequence but then?
> >
> > Redefine it as a bigserial?
>
> That would violate the 32-bit requirement since bigserial would
> make the field a bigint (64 bits). Or am I missing something?
I wasn't sure if that was a limitation he was facing due to business
rules or if he was referring to the limit in postgresql.
Any method that tries to reuse sequence numbers is a bad idea (TM) and
to be avoided, so my point in asking was if the user didn't know about
bigserial / bigint.
From | Date | Subject | |
---|---|---|---|
Next Message | Andrew Sullivan | 2005-01-13 21:43:49 | Re: Column with recycled sequence value |
Previous Message | Michael Fuhr | 2005-01-13 21:19:03 | Re: Column with recycled sequence value |