From: | Michael Nolan <htfoot(at)gmail(dot)com> |
---|---|
To: | Casey Deccio <casey(at)deccio(dot)net> |
Cc: | "pgsql-general(at)postgresql(dot)org" <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: alter column type |
Date: | 2015-06-05 18:36:12 |
Message-ID: | CAOzAqu+55Mbb2Ca2tNfiD6UE1J+kMH4KYCTaKdTqhj1VRm41EQ@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On Fri, Jun 5, 2015 at 12:23 PM, Casey Deccio <casey(at)deccio(dot)net> wrote:
> I have a database in which one table references the primary key of
> another. The type of the primary key was initially int, but I changed it
> to bigint. However, I forgot to update the type of a column that
> references it. So, I've initiated "ALTER TABLE foo ALTER COLUMN bar TYPE
> bigint", where foo/bar is the table/column referencing the primary key that
> is now of type bigint.
>
> However, with 2^31 rows, it is taking a "long" time to write the rows
> (it's been 12 hours). Is there a more efficient way to do this? Even
> if/when this one finishes, there are other column types that I have to
> update. This update effectively locked me out of all access to the data
> anyway, so I don't foresee any concern of writes that might affect
> integrity.
>
> Cheers,
> Casey
>
Probably too late for this time, but in the past when I've needed to
redefine the type for a column, I've made a dump, edited the dump file to
change the type and then renamed the table and reloaded it. That's usually
several orders of magnitude faster.
--
Mike Nolan
nolan(at)tssi(dot)com
From | Date | Subject | |
---|---|---|---|
Next Message | Alvaro Herrera | 2015-06-05 18:36:30 | Re: [GENERAL] 9.4.1 -> 9.4.2 problem: could not access status of transaction 1 |
Previous Message | Tom Lane | 2015-06-05 18:33:12 | Re: [HACKERS] 9.4.1 -> 9.4.2 problem: could not access status of transaction 1 |