| From: | Markus Wanner <markus(at)bluegap(dot)ch> |
|---|---|
| To: | Andres Freund <andres(at)2ndquadrant(dot)com> |
| Cc: | pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: logical changeset generation v3 - comparison to Postgres-R change set format |
| Date: | 2012-11-17 09:22:40 |
| Message-ID: | 50A75760.7030400@bluegap.ch |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On 11/16/2012 03:14 PM, Andres Freund wrote:
> Whats the data type of the "COID" in -R?
It's short for CommitOrderId, a 32bit global transaction identifier,
being wrapped-around, very much like TransactionIds are. (In that sense,
it's global, but unique only for a certain amount of time).
> In the patchset the output plugin has enough data to get the old xid and
> the new xid in the case of updates (not in the case of deletes, but
> thats a small bug and should be fixable with a single line of code), and
> it has enough information to extract the primary key without problems.
It's the xmin of the old tuple that Postgres-R needs to get the COID.
Regards
Markus Wanner
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Michael Paquier | 2012-11-17 10:14:06 | Re: [PATCH 05/14] Add a new relmapper.c function RelationMapFilenodeToOid that acts as a reverse of RelationMapOidToFilenode |
| Previous Message | Markus Wanner | 2012-11-17 09:05:48 | Re: logical changeset generation v3 - comparison to Postgres-R change set format |