The following bug has been logged online:

Bug reference: 2466
Logged by: P.P.S. Narayan
Email address: ppsnarayan (AT) lucent (DOT) com
PostgreSQL version: 8.1
Operating system: Windows XP
Description: lock issues...
Details:

Here is what I am seeing.

I have a postgres instance running on a Windows XP machine with a Tomcat
server. Occasionally one of the http requests (which updates the DB) hangs.
We still have not found out why that happens.

However, there are a bunch of http requests that come in after the hanging
request, which get deadlocked on the locks held by the hanging request. AS
far as I have read, there is a lock_timeout parameter in postgresql config
file to timeout after 1s. But none of the latter http requests timeout.

Can you provide me with
a. Any advice, on how I can go about figuring out what is happening?
b. Is this a known problem on Windows XP. And is there some remedy?

Thanks
-ppsn

---------------------------(end of broadcast)---------------------------
TIP 2: Don't 'kill -9' the postmaster

""P.P.S. Narayan"" <ppsnarayan (AT) lucent (DOT) com> wrote
There is a deadlock_timeout parameter in the configure file, which is "the
time in milliseconds to wait on lock before checking for deadlock." If a
deadlock is resolved, you will see an error message like "deadlock detected"
with some details. Also, you can see current lock status via pg_locks view
to see "information about the locks held by open transactions within the
database server". More details can be found in the docs.

Regards,
Qingqing



---------------------------(end of broadcast)---------------------------
TIP 6: explain analyze is your friend