[OpenAFS] touch: creating `/afs/x': No such file or directory

Torbjorn Pettersson tobbe@strul.nu
12 Jun 2002 23:11:27 +0200


Turbo Fredriksson <turbo@bayour.com> writes:

> With a fools persistency I'm continuing to try to get OpenAFS
> to work on my Sparc at home...
> 

 Change to kernel 2.4.18. There is something b0rken in the
2.2.18(/19?) for sparc+openafs+debian 3.0. I've experienced the
same problem on two sun4m machines. (a sparc5 and a
Classic). Changing to the 2.4.18 kernel helped. Also, this
problem started occuring when I switched to 2.2.18 and updated
afs version at the same time, so changing to lower kernel
version and/or afs version _might_ work, but it is nothing I've
tested. (Only the Classic is mine, and it is not the best
machine to start building new kernels on. think the last build
of the kernel took 14 hours.... :-(
)

> ----- s n i p -----
> tuzjfi:~# kinit turbo && aklog
> Password for turbo@BAYOUR.COM:
> tuzjfi:~# fs mkm /afs/bayour.com root.cell
> fs: File '/afs/bayour.com' doesn't exist
> tuzjfi:~# echo `hostname`.`dnsdomainname`
> tuzjfi.bayour.com
> tuzjfi:~# mount | egrep 'afs|vice'
> /var/lib/openafs/vicepa on /vicepa type ext2 (rw,loop=/dev/loop0)
> AFS on /afs type afs (rw)
> tuzjfi:~# vos listvol tuzjfi
> Total number of volumes on server tuzjfi partition /vicepa: 2
> root.afs                          536870912 RW          2 K On-line
> root.cell                         536870918 RW          2 K On-line
> 
> Total volumes onLine 2 ; Total volumes offLine 0 ; Total busy 0
> 
> tuzjfi:~# ls /afs
> tuzjfi:~# touch /afs/x
> touch: creating `/afs/x': No such file or directory
> tuzjfi:~#
> ----- s n i p -----
> 
> The CellServDB files again (see previous mails about my problems).
> ----- s n i p -----
> tuzjfi:~# cat /etc/openafs/CellServDB
> >bayour.com
> 192.168.1.4 # tuzjfi.bayour.com
> >grand.central.org      #GCO Public CellServDB 10 Apr 2002
> [...]
> tuzjfi:~# head /etc/openafs/server/CellServDB
> >bayour.com     #Cell name
> 192.168.1.4    #tuzjfi
> tuzjfi:~# ifconfig
> eth0      Link encap:Ethernet  HWaddr 08:00:20:71:D3:16
>           inet addr:192.168.1.4  Bcast:192.168.1.255  Mask:255.255.255.0
>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
>           RX packets:2817 errors:0 dropped:0 overruns:0 frame:0
>           TX packets:3509 errors:1 dropped:0 overruns:0 carrier:1
>           collisions:2 txqueuelen:100
>           RX bytes:535013 (522.4 KiB)  TX bytes:354541 (346.2 KiB)
>           Interrupt:38 Base address:0x9400
> 
> eth0:0    Link encap:Ethernet  HWaddr 08:00:20:71:D3:16
>           inet addr:192.168.1.254  Bcast:192.168.1.255  Mask:255.255.255.0
>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
>           Interrupt:38 Base address:0x9400
> [...]
> ----- s n i p -----
> 
> 
> Doing a strace on 'fs', I get this: http://www.bayour.com/strace-fs.out
> 
> There's one part that I don't understand/like. The 'nfssrv()' call.
> This last time I disabled all NFS mounts, NFS server etc. NOTHING
> NFS related is running on tuzjfi (that I know of anyway :).
> 
> I do remember that when I was playing with this on papadoc the first time
> (about 6 months ago), I had problems with the loop device that /vicepa
> was on. Is this problem still there with OpenAFS 1.2.5 (which I'm trying
> to use on tuzjfi)?
> 
> If I try to attach a 'strace' on a 'afsd' process, the strace hangs, not
> giving any output... 
> 
> If I enable either '-verbose' and/or '-debug' it won't mount /afs, and
> the '-debug' causes the 'afsd' to segfault (but processes are still there).
> 
> And just for the record, here's output from 'udebug':
> ----- s n i p -----
> tuzjfi:~# udebug tuzjfi.bayour.com 7002
> Host's addresses are: 192.168.1.4 192.168.1.254
> Host's 192.168.1.4 time is Wed Jun 12 15:39:54 2002
> Local time is Wed Jun 12 15:39:55 2002 (time differential 1 secs)
> Last yes vote for 192.168.1.4 was 0 secs ago (sync site);
> Last vote started 0 secs ago (at Wed Jun 12 15:39:55 2002)
> Local db version is 2.0
> I am sync site forever (1 server)
> Recovery state 1f
> Sync site's db version is 2.0
> 0 locked pages, 0 of them for write
> tuzjfi:~# udebug tuzjfi.bayour.com 7003
> Host's addresses are: 192.168.1.4 192.168.1.254
> Host's 192.168.1.4 time is Wed Jun 12 15:39:57 2002
> Local time is Wed Jun 12 15:39:57 2002 (time differential 0 secs)
> Last yes vote for 192.168.1.4 was 0 secs ago (sync site);
> Last vote started 0 secs ago (at Wed Jun 12 15:39:57 2002)
> Local db version is 1023888363.2
> I am sync site forever (1 server)
> Recovery state 1f
> Sync site's db version is 1023888363.2
> 0 locked pages, 0 of them for write
> Last time a new db version was labelled was:
>          834 secs ago (at Wed Jun 12 15:26:03 2002)
> ----- s n i p -----
> 
> Incase anyone needs to have a look at the logfiles:
> ----- s n i p -----
> ==> BosLog <==
> Wed Jun 12 15:20:58 2002: Server directory access is not okay
> 
> ==> FileLog <==
> Wed Jun 12 15:20:59 2002 File server starting
> Wed Jun 12 15:20:59 2002 afs_krb_get_lrealm failed, using bayour.com.
> Wed Jun 12 15:21:03 2002 VL_RegisterAddrs rpc failed; will retry periodically (code=5376, err=4)
> Wed Jun 12 15:21:04 2002 Partition /vicepa: attached 2 volumes; 0 volumes not attached
> Wed Jun 12 15:21:04 2002 Getting FileServer name...
> Wed Jun 12 15:21:04 2002 FileServer host name is 'tuzjfi'
> Wed Jun 12 15:21:04 2002 Getting FileServer address...
> Wed Jun 12 15:21:04 2002 FileServer tuzjfi has address 192.168.1.4 (0xc0a80104 or 0xc0a80104 in host byte order)
> Wed Jun 12 15:21:04 2002 File Server started Wed Jun 12 15:21:04 2002
> 
> ==> PtLog <==
> Wed Jun 12 15:20:59 2002 Using 192.168.1.4 as my primary address
> 
> ==> VLLog <==
> Wed Jun 12 15:21:00 2002 Using 192.168.1.4 as my primary address
> Wed Jun 12 15:21:01 2002 Starting AFS vlserver 4 (/usr/lib/openafs/vlserver)
> 
> ==> VolserLog <==
> Wed Jun 12 15:21:06 2002 Starting AFS Volserver 2.0 (/usr/lib/openafs/volserver)
> ----- s n i p -----
> 
> The BosLog entry is new though. Previously it's been ok (I had that
> problem on papadoc in the beginning to). From what I can remember
> in conversations with Sam (the Debian GNU/Linux OpenAFS maintainer)
> that don't matter. Checking the BosLog on papadoc, I have the exact
> same error/warning there...
> 
> 
> -- 
> Kennedy PLO quiche kibo iodine Serbian BATF $400 million in gold
> bullion South Africa supercomputer congress spy FBI NSA Saddam Hussein
> [See http://www.aclu.org/echelonwatch/index.html for more about this]
> _______________________________________________
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-info