[OpenAFS] Re: OpenAFS failing rmdir performance

[Derrick Brashear]

Pretty sure either the Solaris VM system paging or Irix's had the original '=
kills the fileserver RPCing in a tight loop' that resulted in this.

Derrick


On Oct 9, 2010, at 4:04 PM, Brandon S Allbery KF8NH <allbery@ece.cmu.edu> wr=
ote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>=20
> On 10/8/10 17:43 , Simon Wilkinson wrote:
>> I do wonder if the abort threshold is too aggressive when applied to auth=
enticated clients. Whilst a denial of service attack is possible from authen=
ticated clients, it's also more likely that there will be administrative mea=
ns available to track down and punish offenders. Maybe we need two different=
 thresholds depending on the class of the client.
>=20
> Wasn't this related to early buggy Windows clients, though?  They were
> authenticated but bringing fileservers to their knees.  It's not uite as
> simple as "we can track them down" when it's screwing your fileserver up i=
n
> production.
>=20
> - --=20
> brandon s. allbery     [linux,solaris,freebsd,perl]      allbery@kf8nh.com=

> system administrator  [openafs,heimdal,too many hats]  allbery@ece.cmu.edu=

> electrical and computer engineering, carnegie mellon university      KF8NH=

> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.10 (Darwin)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
>=20
> iEYEARECAAYFAkywypwACgkQIn7hlCsL25VsNQCfZmVYgHVSsAU3u6Emk1mYftGy
> 9eQAn1qIboqrgW4sECKlLCxrkKoAFjaO
> =3D5gyI
> -----END PGP SIGNATURE-----
> _______________________________________________
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-info