From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Heikki Linnakangas <heikki(at)enterprisedb(dot)com> |
Cc: | alvherre(at)alvh(dot)no-ip(dot)org, Hackers <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: DROP DATABASE and prepared xacts |
Date: | 2007-02-13 03:52:14 |
Message-ID: | 22111.1171338734@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Heikki Linnakangas <heikki(at)enterprisedb(dot)com> writes:
> Tom Lane wrote:
>>> Actually, I think we should completely separate the namespaces of the
>>> global transaction identifiers, so that you could use the same gid in
>>> two different databases without a conflict.
>>
>> Really? They're supposed to be "global".
> Well yeah, the TM should be assigning globally unique ids to every
> transaction. I don't trust all the TM implementations out there, and you
> could even have two different TMs stepping on each others toes, but then
> again I guess it's not really our problem as long as we give a nice
> error message.
If we did that then it'd foreclose the possibility of committing a
prepared xact from a connection in a different DB. Even though I'm a
bit worried about whether we'd have bugs in doing such a thing, I don't
really want to define it to be impossible. "Not implemented" is a lot
different from "impossible because of bad system design".
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2007-02-13 04:19:14 | Re: Variable length varlena headers redux |
Previous Message | Robert Treat | 2007-02-13 03:27:26 | Re: Foreign keys for non-default datatypes, redux |