| From: | Kevin Grittner <kgrittn(at)ymail(dot)com> |
|---|---|
| To: | DT <kurt023(at)hotmail(dot)com>, "pgsql-general(at)postgresql(dot)org" <pgsql-general(at)postgresql(dot)org> |
| Subject: | Re: ALTER TABLE transaction isolation problem |
| Date: | 2013-09-03 14:29:28 |
| Message-ID: | 1378218568.89965.YahooMailNeo@web162905.mail.bf1.yahoo.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
DT <kurt023(at)hotmail(dot)com> wrote:
> I'm reading code of ALTER TABLE, and I found when target table
> needs rewrite, tuple inserted into new heap uses current
> transaction's xid as xmin.
That sure sounds wrong to me.
> Does this behavior satisfy serializable isolation? I wrote some
> test cases:
>
> [ Examples shows that both SERIALIZABLE and REPEATABLE READ
> transactions could see an empty table which was not empty as of
> the point the snapshot was taken. For that matter, it was not
> empty at any later point, either. ]
Why don't we rewrite tuples with their existing xid in such cases?
The current state of affairs seem to me to be a pretty clear bug.
--
Kevin Grittner
EDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
| From | Date | Subject | |
|---|---|---|---|
| Next Message | David Johnston | 2013-09-03 14:42:06 | Re: ALTER TABLE transaction isolation problem |
| Previous Message | Kevin Grittner | 2013-09-03 14:12:54 | Re: SSI and predicate locks - a non-trivial use case |