David wrote:
Will Dormann wrote:
I guess I could try the "ls" trick on the server to see if that
happens to fix anything. Or just go for the 2.6.11.2 kernel as
suggested. (Thanks David!)
I _think_ filesystem corruption was reported around this problem - I'd
upgrade (well, actually, I *did* up
Will Dormann wrote:
I guess I could try the "ls" trick on the server to see if that
happens to fix anything. Or just go for the 2.6.11.2 kernel as
suggested. (Thanks David!)
I _think_ filesystem corruption was reported around this problem - I'd
upgrade (well, actually, I *did* upgrade!)
David
Dan Wilga wrote:
Are you sure NFS statd and lockd are running on the machine serving the
NFS volume? NFS is known for running almost-normally when either of
these is off for some reason.
rpc.statd is running on the NFS server. I'm not so sure about lockd.
"ps aux" lists "[lockd]".That's w
At 9:08 AM +0100 3/31/05, David wrote:
I was surprised - I'm not sure how deleting a file twice could cause NFS
stale file handles...
I agree with David on this. Stale handle errors should only occur if
there is a problem at the NFS level.
Are you sure NFS statd and lockd are running on the machi
Will Dormann wrote:
My NFS server is running 2.6.9 with NFSv3
The filesystem for my store is indeed xfs.
In that case... upgrade the kernel to 2.6.11.2 (no later)
Now :)
There were quite a few issues around that time to do with xfs/nfs.
search lkml for "XFS: inode with st_mode == 0" (though that we
David wrote:
I was surprised - I'm not sure how deleting a file twice could cause NFS
stale file handles...
I'd suggest the lirc/delete could be a mis-diagnosis on your part Will,
I think that this could be more of an NFS problem.
Perhaps. I'm still not convinced that it's *not* the problem, but
Chris Pinkham wrote:
Chris Pinkham wrote:
Thanks for the good descriptiong. I'll try to simulate here and get a
fix into CVS if I can find an issue. I think I may already know what
the issue is and how to prevent this.
Excellent! Let me know what you find out. I'm currently runnin
> Chris Pinkham wrote:
> > Thanks for the good descriptiong. I'll try to simulate here and get a
> > fix into CVS if I can find an issue. I think I may already know what
> > the issue is and how to prevent this.
>
> Excellent! Let me know what you find out. I'm currently running
> 0.16, so
Chris Pinkham wrote:
Thanks for the good descriptiong. I'll try to simulate here and get a
fix into CVS if I can find an issue. I think I may already know what
the issue is and how to prevent this.
Excellent! Let me know what you find out. I'm currently running
0.16, so hopefully it's someth
> Anyway, what I'm thinking is happening is this...
>
> When I go to delete a recording, there are *two* "Enter" presses
> registered in rapid succession. (Due to the flakey joystick). The
> first one goes and deletes the file from my NFS store, but mythtv goes
> to delete the file a second
Some time ago I posted a message about a problem with MythTV spewing out
"Stale NFS file handle" messages. It seemed to correlate with my
upgrade to MythTV 0.17, so I figured that was the cause.
After reverting to 0.16, it seemed to be working well, but I recently
ran into it a few times. I
11 matches
Mail list logo
