Re: novfsd causes heavy CPU load on SLED11

[W_ Prindl]

Is this on a real or virtual machine? Unfortunately on some cpu types,
this behaviour is often occurring and more likely on virtual, than on
real machines. Besides of this are you sure there is no beagle indexer
or anything similar running over your NCP connected directories.

--
W. Prindl

redwood wrote:

>
>Hi folks,
>
>I recently installed a SLED11 (64-Bit) on my notebook. The SLED11 is
>fully patched.
>
>I installed the Novell Client 2.0 SP2 for Linux and it is working. The
>Login, the tray
>everything is perfect.
>
>During the last days I noticed that the CPU load is quite heavy. I
>found out that the novfsd is doing a lot and keeps the CPU running on
>100%. After a service-restart everything is fine for a while.
>
>Do you have any idea about that?
>
>
>Code:
>--------------------
> adrian@asc:~> uname -r
> 2.6.27.29-0.1-default
>
>--------------------
>
>
>
>Code:
>--------------------
> novell-client-script-2.0.0-32
> yast2-novell-client-2.0.0-22
> novell-client-2.0.0-29
>
>--------------------
>
>
>Thanks,
>Adrian

[W_ Prindl]

In this configuration I did not see that behaviour, but I am still at
32-bit SLED11. Maybe the 64-bit client has some issues - but that's
just a wild guess.
--
W. Prindl

redwood wrote:

>
>Hi,
>
>the Beagle Indexer is not running. It is a physical machine running on
>a hp 6910p Notebook.
>
>BR,
>Adrian

[a...@novell.com]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I am also running SLED 11 x86_64 (no patches though) with the Novell
Client and do not see this, but I also do not run with drive mappings all
day, every day. In the past month or so I have some fairly heavy load
testing with drives mapped (lots of operations, not lots of data) and in
those cases I have seen novfsd working fairly hard, but I really am
pushing that service at the time (while loop touching files simply to
create them and, immediately after, delete them as fast as possible;
auditing testing on NetWare is the purpose). If you disconnect drive
mappings does the problem continue?

Good luck.

W_ Prindl wrote:
> In this configuration I did not see that behaviour, but I am still at
> 32-bit SLED11. Maybe the 64-bit client has some issues - but that's
> just a wild guess.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
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=2ECe
-----END PGP SIGNATURE-----