Bug 689308 - fedora 14 nfs client stops working randomly - kernel backtrace raw_local_irq_restore _raw_spin_unlock_irqrestore
Summary: fedora 14 nfs client stops working randomly - kernel backtrace raw_local_irq_...
Keywords:
Status: CLOSED INSUFFICIENT_DATA
Alias: None
Product: Fedora
Classification: Fedora
Component: kernel
Version: 14
Hardware: x86_64
OS: Linux
unspecified
high
Target Milestone: ---
Assignee: Kernel Maintainer List
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2011-03-21 03:03 UTC by Edgar Hoch
Modified: 2011-10-11 20:01 UTC (History)
7 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2011-10-11 20:01:16 UTC
Type: ---


Attachments (Terms of Use)
Kernel messages from /var/log/messages (18.74 KB, text/plain)
2011-03-21 03:03 UTC, Edgar Hoch
no flags Details
Kernel backtrace messages from /var/log/messages from another host (85.98 KB, text/plain)
2011-03-21 16:32 UTC, Edgar Hoch
no flags Details

Description Edgar Hoch 2011-03-21 03:03:38 UTC
Created attachment 486533 [details]
Kernel messages from /var/log/messages

Description of problem:
We use fedora 14 x86_64 on servers and pcs which mount nfs file systems from other servers. It happens from time to time that a server or pc, which are nfs clients, stops working. Then we cannot login via ssh and cannot do local work - any access to a nfs mounted directory "hangs" - waiting to recieve the nfs data forever.

The problem occurs on fedora 14 nfs clients, it occurs randomly - sometimes all works fine and later all nfs access either hangs or get an i/o error. As I can see using tcpdump, strace etc. the nfs access doesn't create nfs network traffic when the error occurs - so I think the problem is in the nfs client - in the kernel as this code is part of the kernel. It occurs with nfs v3 and v4 and with different nfs servers and different nfs server operating systems.

Most of the time we got not error messages when the computer is not working any more. But now I found kernel backtraces in /var/log/messages after I had hard resetted the server (a nfs client). So I hope this will help to solve the problem when I report this kernel messages.


Version-Release number of selected component (if applicable):
kernel-2.6.35.11-83.fc14.x86_64
nfs-utils-1.2.3-5.fc14.x86_64
nfs-utils-lib-1.1.5-3.fc14.x86_64
nfs4-acl-tools-0.3.3-6.fc13.x86_64


How reproducible:
It occurs randomly. I don't know how to reproduce - it happens sometimes and sometimes not.

  
Actual results:
Fedora 14 system with nfs client (for homes, /usr/local, /mount/* etc. automouted) sometimes stops working. Only hard reboot via hardware reset helps.

Expected results:
Fedora 14 system with nfs client (for homes, /usr/local, /mount/* etc. automouted) should run without failure.

Comment 1 Edgar Hoch 2011-03-21 16:32:11 UTC
Created attachment 486637 [details]
Kernel backtrace messages from /var/log/messages from another host

Attached is another kernel backtrace from another host. It seems also related to nfs. Is there a deadlock in the kernel?

Comment 2 Josh Boyer 2011-08-29 15:36:22 UTC
Are you still seeing this with the latest f14 kernel?

Comment 3 Dave Jones 2011-10-11 20:01:16 UTC
at this stage in its lifecycle, f14 isn't going to get any significant NFS changes other than security related patches.  If you can reproduce this problem with f15 or newer, please reopen.


Note You need to log in before you can comment on or make changes to this bug.