Re: Problem with ALTER TABLE - occasional "tuple concurrently updated"

On Thu, Mar 10, 2011 at 10:37 PM, Bruce Momjian <bruce(at)momjian(dot)us> wrote:
> Robert Haas wrote:
>> On Thu, Mar 10, 2011 at 4:08 PM, Bruce Momjian <bruce(at)momjian(dot)us> wrote:
>> > Was this fixed?
>>
>> Not yet.  I can probably fix it, if nobody else wants to do it.
>
> Well, it has languished for five months, so the "nobody else wants" part
> is probably accurate.  ;-)

OK. Do we want to back-patch this, and if so how far? On the one
hand, the symptom that OP is experiencing clearly sucks for him, but
on the other hand upgrading the strength of a lock in releases that
have been out in the field for a long time seems like an open
invitation to have the villagers show up with pitchforks. Then again,
ShareUpdateExclusiveLock doesn't interfere with routine queries, so
maybe it's no big deal. Given that we have only one report, I'm
inclined to just fix it in the master branch, but I could easily be
talked into the other approach if someone wants to make an argument
for it.

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