From: christoph.beyer on
Hi,

this is an old problem it was worse in the pre-3.5.0 releases, on solaris
10 all the cpu gets consumed by fcntl calls:

7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFE050) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDE78) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF58) = 0
7963: fcntl(22, F_SETLKW64, 0xFFBFDF70) = 0

[ snip ]

It happens during every printing related transaction (don't do
fileservice, probably no difference) in the pre 3.5.0 releases it happened
already when the client only requested the print queue state. Now it
happens when any printing transactions are in progress. It slows down the
machine completely and makes the service unusable. Same setup, same tdb's
work OK on linux.

Seems to be an old problem:

http://lists-archives.org/samba/21962-rogue-process.html

Any chance to fix this, do you need more debug output ?

Can't see much tdb activity with tdb:10 (?)

cheers
~christoph


--
/* Christoph Beyer | Office: Building 2b / 23 *\
* DESY | Phone: 040-8998-2317 *
* - IT - | Fax: 040-8998-4060 *
\* 22603 Hamburg | http://www.desy.de */


--
To unsubscribe from this list go to the following URL and read the
instructions: https://lists.samba.org/mailman/options/samba