On Nov 12, 2007 11:01 AM, Heikki Linnakangas <heikki(at)enterprisedb(dot)com> wrote:
>
> Scott Marlowe wrote:
> > So, between the first and second vacuum you had a long running
> > transaction that finally ended and let you clean up the dead rows.
>
> No, before 8.3, CLUSTER throws away non-removable dead tuples. So the
> long running transaction might still be there.
Wow, good to know. Why would it have changed in 8.3? Was it
considered broken behaviour?