[OpenAFS] Openafs 1.4.8 on OSX very slow

[david l goodrich]

--wac7ysb48OaltWcw
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Sat, Mar 07, 2009 at 10:40:19PM +0100, Hans Melgers wrote:
> I did install the safari 4 beta a few days ago, maybe there's a link ? =
=20
> uninstalled it again but no improvement. Anybody else running safari 4 =
=20
> beta without having problems in afs ?

I'm running the Safari 4 beta and WebKit nightlies as my primary           =
    =20
browsers, and have no AFS problems on my Macbook, running 10.5.6.          =
    =20
  --david

> On 6 mrt 2009, at 17:25, Joel wrote:
>
>>
>> Does dmesg report any issues with contacting any servers?
>>
>> On Thu, Mar 5, 2009 at 9:10 PM, Derrick Brashear <shadow@gmail.com> =20
>> wrote:
>>> On Thu, Mar 5, 2009 at 8:10 PM, Jason Edgecombe=20
>>> <jason@rampaginggeek.com> wrote:
>>>> Hans Melgers wrote:
>>>>>
>>>>> I notice afs performance on osx Leopard is very slow; opening a =20
>>>>> folder in
>>>>> Finder takes up to 2 minutes to even display. I have the right =20
>>>>> tokens,
>>>>> firewall is open for all connections.
>>>>> Performance in terminal is much better, a cd and ls to /afs works
>>>>> instantly. So it seems there are problems in the finder.
>>>>> I noticed two afsd processes running. normal ?
>>>>>
>>>>> Im also tried running parallels with the windows client on the same
>>>>> machine, works perfect and very fast.
>>>>>
>>>>> Does anyone have some tips to speed up performance ? Are there =20
>>>>> any known
>>>>> problems in the mac osx finder in the latest leopard?
>>>>> I'm running 1.4.8 with openafsconfig, MIT kerberos extras. Tokens=20
>>>>> look
>>>>> fine. Mac OSX 10.5.6
>>>>
>>>> Was /afs the folder that you tried to view in finder? If so, then =20
>>>> it will be
>>>> extremely slow with finder because finder will try to stat all of =20
>>>> the cells,
>>>> which will take a while. Try a deeper folder, like
>>>> /afs/grand.central.org/software/ and see if it acts better.
>>>
>>> with -fakestat-all that shouldn't be true. there is the issue that
>>> bulkstat does not (currently) work, which will make things slower, =20
>>> but
>>> it shouldn't be horrible.
>>>
>>> --
>>> Derrick
>>> _______________________________________________
>>> OpenAFS-info mailing list
>>> OpenAFS-info@openafs.org
>>> https://lists.openafs.org/mailman/listinfo/openafs-info
>>>
>> _______________________________________________
>> OpenAFS-info mailing list
>> OpenAFS-info@openafs.org
>> https://lists.openafs.org/mailman/listinfo/openafs-info
>
> Hans Melgers
> ENEMB.V.
>
> www.enem.nl
> hans@enem.nl
>
>
>
>
> _______________________________________________
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-info

--wac7ysb48OaltWcw
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAkmy7dsACgkQHDmo5jqnP4QQAgCdHe/1tfpUbfhzSEZfOCOeZJYF
j8oAn1PI3/0/gfNEI8Oo+7vc8E3h5pAo
=U/4Z
-----END PGP SIGNATURE-----

--wac7ysb48OaltWcw--