Re: Commit hung bug

S H wrote:
> There is one bug mentioned commit hung for days..
> http://www.postgresql.org/message-id/1AF3044FCAB26F4DB1AE551F8A33634B3D22FB@mail.digital-rapids.com > The interesting thing would be to see the server logs, not the> application logs. Specifically, an issue that could look just likethis> was fixed in 8.1.7, in which case you would see weird error messages> about permission denied or such in the *server* logs. None of thatwould> show up in the client logs.
> Any idea what exactly is this bug.
> I could not make out relation between release notes mentioned in http://www.postgresql.org/docs/8.1/static/release-8-1-7.html
> and above comment.

Maybe it's this commit, which was part of 8.1.6:

commit 9f1b531420ee13d04c7701b34bb4b874df7ff2fa
Author: Teodor Sigaev <teodor(at)sigaev(dot)ru>
Date: Fri Oct 13 14:00:17 2006 +0000

Fix infinite sleep and failes of send in Win32.

1) pgwin32_waitforsinglesocket(): WaitForMultipleObjectsEx now called with
finite timeout (100ms) in case of FP_WRITE and UDP socket. If timeout occurs
then pgwin32_waitforsinglesocket() tries to write empty packet goes to
WaitForMultipleObjectsEx again.

2) pgwin32_send(): add loop around WSASend and pgwin32_waitforsinglesocket().
The reason is: for overlapped socket, 'ok' result from
pgwin32_waitforsinglesocket() isn't guarantee that socket is still free,
it can become busy again and following WSASend call will fail with
WSAEWOULDBLOCK error.

See http://archives.postgresql.org/pgsql-hackers/2006-10/msg00561.php

It's troubling to be talking about a bug that was patched in 2006 for
the 8.1.6 release, however. Anything prior to that is not something
anyone should be using anymore. At the very least, you should have
migrated to 8.1.23; but 8.1 has been unsupported altogether for more
than two years now. Even 8.2 is out of support.

--
Álvaro Herrera http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services