Hi,

You are right indeed. I had a special case lurking deep in my code
that made a write query on the muxes table. Of course, my stubborn
head was certain that I had no such thing and hence I used a lot of
time chasing ghosts :)

Is it possible to detect that you are waiting for locks to improve the
error message (I guess not, the c library probably just sits there
until it gets the lock)? I'd probably figured this by myself with a
better error message. I guess such an error message should be given by
postgreSQL.. the "FATAL:  08P01: invalid frontend message type 32"
isn't exactly helpful :)

Anyway, the problem is solved and I'm happy.

Thanks a lot,
Frederik

On 10/23/06, Christoph Zwerschke <[EMAIL PROTECTED]> wrote:
> Hi Frederik,
>
> this sounds more like your "outer" process has a transaction running on
> the table "muxes" that is locking the table from updates, so the "inner"
> process waits for ever. What happens when you issue a "commit" before
> starting the inner process?
>
> -- Christoph
>
>
>
> _______________________________________________
> PyGreSQL mailing list
> [email protected]
> http://mailman.vex.net/mailman/listinfo/pygresql
>


-- 
Time was invented by carbon based lifeforms in order to monitor their
ongoing decay
_______________________________________________
PyGreSQL mailing list
[email protected]
http://mailman.vex.net/mailman/listinfo/pygresql

Reply via email to