Bug#489613: openafs-modules-source: attempt to reinsert module results in Cannot allocate memory

2010-08-09 Thread Conrad Hughes

FWIW, my Debian stable machine just started exhibiting this problem
after months of perfect operation; persisted across reboots too.
Switching to the 1.4.12.1+dfsg-2~bpo50+1 openafs suite in
lenny-backports appears to have resolved it.

uname -a:
  Linux ... 2.6.26-2-686 #1 SMP Mon Jun 21 05:58:44 UTC 2010 i686 GNU/Linux

Thanks for the suggestion!

Conrad

-- 
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.




-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Bug#489613: openafs-modules-source: attempt to reinsert module results in Cannot allocate memory

2009-08-21 Thread Russ Allbery
Hi folks,

Several of you reported this problem with reinserting the OpenAFS module
around the 1.4.7 version that shipped with lenny.  If you're still having
trouble with this, could you try the 1.4.11 version that's available from
backports.org?  I think the problem may be solved.  I was able to
reproduce this problem on an amd64 system before, but I can't reproduce it
any more with the current module.

-- 
Russ Allbery (r...@debian.org)   http://www.eyrie.org/~eagle/



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Bug#489613: openafs-modules-source: attempt to reinsert module results in Cannot allocate memory

2009-02-16 Thread Zach JL
Package: openafs-modules-source
Version: 1.4.7.dfsg1-6
Severity: higher than normal
Justification: Package is unusable

I have the same issue on lenny.  Steps take were:

apt-get install module-assistant
m-a a-i openafs

When trying to modprobe the openafs module, I get this in dmes:

[87485.390787] kmem_cache_create: duplicate cache afs_inode_cache
[87485.394833] Pid: 32096, comm: modprobe Tainted: P 
2.6.26-1-amd64 #1
[87485.399242]
[87485.399243] Call Trace:
[87485.407689]  [80251ca7] get_online_cpus+0x22/0x38
[87485.412005]  [80297c8d] kmem_cache_create+0x407/0x44a
[87485.416287]  [a046ad23] :openafs:init_once+0x0/0x8
[87485.420471]  [8024096d] groups_alloc+0x3c/0xc5
[87485.424629]  [a046ad11] :openafs:afs_init_inodecache+0x23/0x35
[87485.428923]  [a002702b] :openafs:init_module+0x2b/0xd5
[87485.433156]  [80255067] sys_init_module+0x190e/0x1aa4
[87485.437371]  [802441db] find_task_by_vpid+0x0/0x1d
[87485.441597]  [80224da4] cstar_do_call+0x1b/0x65



and this at stdout:

FATAL: Error inserting openafs
(/lib/modules/2.6.26-1-amd64/fs/openafs.ko): Cannot allocate
memory

uname -r
2.6.26-1-amd64





-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Bug#489613: openafs-modules-source: attempt to reinsert module results in Cannot allocate memory

2009-02-16 Thread Russ Allbery
Zach JL xaque...@gmail.com writes:

 Package: openafs-modules-source
 Version: 1.4.7.dfsg1-6
 Severity: higher than normal
 Justification: Package is unusable

 I have the same issue on lenny.  Steps take were:

 apt-get install module-assistant
 m-a a-i openafs

 When trying to modprobe the openafs module, I get this in dmes:

 [87485.390787] kmem_cache_create: duplicate cache afs_inode_cache
 [87485.394833] Pid: 32096, comm: modprobe Tainted: P 
 2.6.26-1-amd64 #1

You get this despite not having ever had any openafs module ever loaded on
that system prior?

Again, I can't duplicate this problem on any system that I have, so I'm
not sure what's causing it or how to fix it.  Just to make sure, you're
not loading the kafs module that comes with the kernel for some reason,
are you?

-- 
Russ Allbery (r...@debian.org)   http://www.eyrie.org/~eagle/



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Bug#489613: openafs-modules-source: attempt to reinsert module results in Cannot allocate memory

2008-07-08 Thread Andrew J Perrin
When I don't reload openafs-client and vpnc drops and restarts, afs just 
hangs when I try to cd into /afs or a subdirectory thereof.


--
Andrew J Perrin - [EMAIL PROTECTED] - http://perrin.socsci.unc.edu
Associate Professor of Sociology; Book Review Editor, _Social Forces_
University of North Carolina - CB#3210, Chapel Hill, NC 27599-3210 USA


On Mon, 7 Jul 2008, Russ Allbery wrote:


Andrew J Perrin [EMAIL PROTECTED] writes:


Probably quite a few. My setup is such that AFS only works to my
institution when a VPN connection is in place, and the VPN connection
has been unstable recently. When the VPN connection drops, I have to do:

/etc/init.d/openafs-client stop
vpnc --dpd-idle 0
/etc/init.d/openafs-client start


Hm, well, that's a problem in and of itself.  Why do you have to disable
AFS when your VPN connection isn't available?  AFS should cope with the
VPN connection going away and coming back.  What happens if you don't do
this?

--
Russ Allbery ([EMAIL PROTECTED])   http://www.eyrie.org/~eagle/





--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Bug#489613: openafs-modules-source: attempt to reinsert module results in Cannot allocate memory

2008-07-08 Thread Russ Allbery
Andrew J Perrin [EMAIL PROTECTED] writes:

 When I don't reload openafs-client and vpnc drops and restarts, afs just
 hangs when I try to cd into /afs or a subdirectory thereof.

How long have you waited on AFS when this happens?  Talking this over with
upstream, the general consensus is that it shouldn't take longer than a
few minutes for AFS to figure it out; if it takes longer than ten minutes,
there's definitely some other bug here that we'd like to get to the bottom
of.

(That may still be too slow, I realize, but I'd like to get a handle on
whether AFS is just being too slow to realize that its IP address has
changed or whether it's really getting completely stuck.  We may be able
to fix both problems, but need to understand first which problem we're
dealing with.)

BTW, this bug is almost certainly present in earlier versions of AFS as
well; AFS has leaked memory in one way or another for quite a while on
module reload.  Only recently has it gotten a facility for realizing that
it's done so.

-- 
Russ Allbery ([EMAIL PROTECTED])   http://www.eyrie.org/~eagle/



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Bug#489613: openafs-modules-source: attempt to reinsert module results in Cannot allocate memory

2008-07-07 Thread Andrew J Perrin

On Sun, 6 Jul 2008, Russ Allbery wrote:


Andrew Perrin [EMAIL PROTECTED] writes:


Package: openafs-modules-source
Version: 1.4.7.dfsg1-2
Severity: grave
Justification: renders package unusable


I think that's a little strong.  I downgraded the bug.



OK by me :)


adorno:/etc/openafs# /etc/init.d/openafs-client force-reload
Stopping AFS services:afsd: Shutting down all afs processes and afs state
 openafs.
Starting AFS services:FATAL: Error inserting openafs 
(/lib/modules/2.6.24/fs/openafs.ko): Cannot allocate memory

Failed to load AFS kernel module, not starting AFS
adorno:/etc/openafs# modprobe openafs
FATAL: Error inserting openafs (/lib/modules/2.6.24/fs/openafs.ko): Cannot 
allocate memory


How many times have you unloaded and reloaded the module on that system
since the last reboot?


Probably quite a few. My setup is such that AFS only works to my 
institution when a VPN connection is in place, and the VPN connection has 
been unstable recently. When the VPN connection drops, I have to do:


/etc/init.d/openafs-client stop
vpnc --dpd-idle 0
/etc/init.d/openafs-client start


...
If you have more useful debugging information (dmesg, for example), that
will be helpful; I can provide upstream with the details of the memory
leak that I see, but I can reload the module anyway.



Thanks.

--
Andrew J Perrin - [EMAIL PROTECTED] - http://perrin.socsci.unc.edu
Associate Professor of Sociology; Book Review Editor, _Social Forces_
University of North Carolina - CB#3210, Chapel Hill, NC 27599-3210 USA






--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Bug#489613: openafs-modules-source: attempt to reinsert module results in Cannot allocate memory

2008-07-07 Thread Russ Allbery
Andrew J Perrin [EMAIL PROTECTED] writes:

 Probably quite a few. My setup is such that AFS only works to my
 institution when a VPN connection is in place, and the VPN connection
 has been unstable recently. When the VPN connection drops, I have to do:

 /etc/init.d/openafs-client stop
 vpnc --dpd-idle 0
 /etc/init.d/openafs-client start

Hm, well, that's a problem in and of itself.  Why do you have to disable
AFS when your VPN connection isn't available?  AFS should cope with the
VPN connection going away and coming back.  What happens if you don't do
this?

-- 
Russ Allbery ([EMAIL PROTECTED])   http://www.eyrie.org/~eagle/



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Bug#489613: openafs-modules-source: attempt to reinsert module results in Cannot allocate memory

2008-07-06 Thread Andrew Perrin
Package: openafs-modules-source
Version: 1.4.7.dfsg1-2
Severity: grave
Justification: renders package unusable



-- System Information:
Debian Release: lenny/sid
  APT prefers testing
  APT policy: (500, 'testing')
Architecture: i386 (i686)

Kernel: Linux 2.6.24 (SMP w/4 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/bash

Versions of packages openafs-modules-source depends on:
ii  bison   1:2.3.dfsg-5 A parser generator that is compati
ii  debhelper   7.0.10   helper programs for debian/rules
ii  flex2.5.35-2 A fast lexical analyzer generator.
ii  kernel-package  11.001-0.1   A utility for building Linux kerne
ii  module-assistant0.10.11.0tool to make module package creati

openafs-modules-source recommends no packages.

-- no debconf information

adorno:/etc/openafs# /etc/init.d/openafs-client force-reload
Stopping AFS services:afsd: Shutting down all afs processes and afs state
 openafs.
Starting AFS services:FATAL: Error inserting openafs 
(/lib/modules/2.6.24/fs/openafs.ko): Cannot allocate memory

Failed to load AFS kernel module, not starting AFS
adorno:/etc/openafs# modprobe openafs
FATAL: Error inserting openafs (/lib/modules/2.6.24/fs/openafs.ko): Cannot 
allocate memory



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Bug#489613: openafs-modules-source: attempt to reinsert module results in Cannot allocate memory

2008-07-06 Thread Russ Allbery
Andrew Perrin [EMAIL PROTECTED] writes:

 Package: openafs-modules-source
 Version: 1.4.7.dfsg1-2
 Severity: grave
 Justification: renders package unusable

I think that's a little strong.  I downgraded the bug.

 adorno:/etc/openafs# /etc/init.d/openafs-client force-reload
 Stopping AFS services:afsd: Shutting down all afs processes and afs state
  openafs.
 Starting AFS services:FATAL: Error inserting openafs 
 (/lib/modules/2.6.24/fs/openafs.ko): Cannot allocate memory
 
 Failed to load AFS kernel module, not starting AFS
 adorno:/etc/openafs# modprobe openafs
 FATAL: Error inserting openafs (/lib/modules/2.6.24/fs/openafs.ko): Cannot 
 allocate memory

How many times have you unloaded and reloaded the module on that system
since the last reboot?  I can't duplicate this on any of my systems at the
moment, but I have seen this after some period of time of testing various
versions without a reboot and the AFS module does tend to leak at least
some memory when it's unloaded.  I suspect that eventually it fragments
the kernel's internal memory allocator.  (The Linux kernel interface keeps
changing and it's very hard to keep up with the changes in areas like
this.)

I suspect the system will be fine after a reboot.

Unloading and reloading the module *mostly* works but is not the best
thing to do when it can be avoided.  It's not something that I consider
critical to support, although I'll report this upstream and hopefully
someone will have a chance to take a look and figure out where we're
leaking memory.

If you have more useful debugging information (dmesg, for example), that
will be helpful; I can provide upstream with the details of the memory
leak that I see, but I can reload the module anyway.

-- 
Russ Allbery ([EMAIL PROTECTED])   http://www.eyrie.org/~eagle/



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]