On Jul 7, 2004, at 8:11 PM, Christine Nattrass wrote:
Ok, answers to miscellaneous questions
I am running Debian testing (2.6.7 kernel) filesystem for cache - XFS (how do I change this?)
Create another partition and make an ext2 filesystem on that. Then mount it as your AFS cache.
platform - IA32, Dell laptop: Inspiron 5100 with a pentium 4 processor
here is the info about the patch I was using http://linux.ncsu.edu/projects/openafs-rpms/
It appears I was having some problems getting the kernel module to load properly. That was one of the problems.
Here's what I do when I try to run openafs now:
[EMAIL PROTECTED]:~/skype_ver-0_90_0_6$ su
Password:
tesla:/home/madscientistgirl/skype_ver-0_90_0_6# insmod /usr/local/lib/openafs/libafs-2.6.7.ko
tesla:/home/madscientistgirl/skype_ver-0_90_0_6# dmesg
...
libafs: module license 'unspecified' taints kernel.
Found sys_call_table at c0442200
tesla:/usr/local# /usr/local/sbin/afsd -memcache
afsd: WARNING: Cache dir check failed (unable to stat cache base directory)
afsd: All AFS daemons started.
afsd: Can't mount AFS on /afs(22)
tesla:/usr/local#
Is it the same patch you used last time? I'm asking because Jack Neely changed things.
...and is your cell working?? This means you have the ability to use the Cell from another client, other than your Linux box.
Do I have hyperthreading enabled? I am working on finding out but I'm having trouble checking. I'm still relatively new to Linux. But I'm no longer getting a segmentation fault.
a simple way is a look into the proc filesystem if you see two cpus it's a good chance that it's enabled
I evidently do have it set up so that was trying to use an xfs file system for the cache -- how do I change this? Why won't xfs and work? Is there any more information on this?
Yes, ext2 and ext3 are working because AFS supports them :-)
If the error persisits after you change the cache to ext2 you should start the client (afsd) with -verbose -debug and post the output.
As the client said, it can't mount AFS. ;-)
From the information above I can't tell why.
Horst
_______________________________________________ OpenAFS-devel mailing list [EMAIL PROTECTED] https://lists.openafs.org/mailman/listinfo/openafs-devel
