Re: Set new system identifier using pg_resetxlog

On Tue, Jul 1, 2014 at 11:19 AM, Andres Freund <andres(at)2ndquadrant(dot)com> wrote:
>> I am, however, kind of frustrated, still, that the pg_computemaxlsn
>> patch, which I thought was rather a good idea, was scuttled by the
>> essentially that same objection: let's not extend pg_resetxlog &
>> friends because people might use the new functionality to do bad
>> things and then blame us.
>
> Well, the reasons were a bit different. Senior community members
> repeatedly suggested that it'd be usable for faillback - and it's not a
> unreasonable to think it is. Even though it'd fail subtly because of
> hint bit and related reasons.
> In contrast you have to be pretty desperate to think that you could make
> two clusters replicate from each other by just fudging pg_control long
> enough, even if the clusters aren't actually related.

Well, I still think it's pretty likely someone could make that mistake
here, too. Maybe not a senior community member, but somebody.
However, I think the right answer in that case and this one is to tell
the person who has made the mistake "you screwed up" and move on.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company