[OpenAFS] No buffer space available

[Stephan Wiesand]

On Jul 23, 2013, at 18:40 , Andrew Deason <adeason@sinenomine.net> =
wrote:

> On Tue, 23 Jul 2013 13:11:17 +0200
> Hans-Werner Paulsen <hans@MPA-Garching.MPG.DE> wrote:
>=20
>> sometimes creating a file (using different programs) fails with the
>> error message "No buffer space available". This is on amd64_linux26
>> with OpenAFS 1.6.2 (both client and server). Any idea?
>=20
> I don't see anywhere we'd be generating that error code (ENOBUFS), and =
I
> can't see how it would show up that way if we got it back from a =
socket
> or something. Do you have any idea if the machine is under a lot of =
load
> or memory pressure, or if the directory has a lot of entries in it?


To add a data point:

I just encountered this while copying two rather small files for the =
1.6.5.1 release from here (~ Berlin) to PENN-CENTRAL.SRV.CS.CMU.EDU.

% cp /tmp/openafs-release-* . -iv  =20
`/tmp/openafs-release-fedora-1.6.5.1-1.noarch.rpm' -> =
`./openafs-release-fedora-1.6.5.1-1.noarch.rpm'
cp: cannot create regular file =
`./openafs-release-fedora-1.6.5.1-1.noarch.rpm': No buffer space =
available
`/tmp/openafs-release-rhel-1.6.5.1-1.noarch.rpm' -> =
`./openafs-release-rhel-1.6.5.1-1.noarch.rpm'

So the first one failed. The destination file was created, but had =
length 0. The very next attempt succeeded. It's the first time I ever =
encountered this problem. Client is SL6.4, 2.6.32-358.18.1.el6.x86_64, =
OpenAFS 1.6.5.1 (built against the -358.el6 kernel - which I know voids =
my warranty ;-), disk cache on ext4.

Not reproducible, nothing in the logs or the kernel message buffer.

The client was not under any load and has 12 GB of free (really unused) =
RAM. The server seemed not to be busy either. The only special condition =
is the long distance between client and server.

No showstopper, and probably very hard to track down without a =
reproducer.


--=20
Stephan Wiesand
DESY - DV -
Platanenallee 6
15732 Zeuthen, Germany