[OpenAFS] vos listaddrs and "fake" fileserver addresses
Ryan C. Underwood
nemesis-lists@icequake.net
Thu, 10 Sep 2009 18:00:35 -0500
--wRRV7LY7NUeQGEoC
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Thu, Sep 10, 2009 at 06:19:09PM -0400, Derrick Brashear wrote:
>=20
> after adding to the netinfo files, restart *only the fileserver* on
> the hosts. not the vlserver, not all. does anything change? what's in
> the FileLogs after you do?
Way too late to not restart all :-) Just as a note, I've also moved the
sysid files and let them be regenerated in a desperate attempt to get
this working. You can see the results in the below logs.
Let me post some logs with real IP addresses, since I'm at home now.
Also, I erroneously posted the location of the NetInfo file, on Debian
they've been moved to /var/lib/openafs/local (at least according to the
man page).
This is on fileserver 1 (valhalla)
# cat /var/lib/openafs/local/NetInfo=20
10.0.1.232
f 65.38.17.158
# udebug xanadu 7002
Host's addresses are: 10.0.1.230 65.38.17.159=20
Host's 10.0.1.230 time is Thu Sep 10 17:53:41 2009
Local time is Thu Sep 10 17:53:43 2009 (time differential 2 secs)
Last yes vote for 230.1.0.10 was 6 secs ago (sync site);=20
Last vote started 6 secs ago (at Thu Sep 10 17:53:37 2009)
Local db version is 1251646288.3
I am sync site until 51 secs from now (at Thu Sep 10 17:54:34 2009) (2 se=
rvers)
Recovery state 1f
Sync site's db version is 1251646288.3
0 locked pages, 0 of them for write
Server (10.0.1.232 65.38.17.158): (db 1251646288.3)
last vote rcvd 9 secs ago (at Thu Sep 10 17:53:34 2009),
last beacon sent 6 secs ago (at Thu Sep 10 17:53:37 2009), last vote =
was yes
dbcurrent=3D1, up=3D1 beaconSince=3D1
# vos listaddrs
xanadu.icequake.net
valhalla.icequake.net
# cat PtLog
Thu Sep 10 17:45:11 2009 Using 10.0.1.232 as my primary address
Thu Sep 10 17:45:20 2009 ubik: A Remote Server has addresses: Thu Sep 10 =
17:45:20 2009 10.0.1.230 Thu Sep 10 17:45:20 2009 65.38.17.159 Thu Sep 10 1=
7:45:20 2009=20
# cat FileLog
Thu Sep 10 17:45:11 2009 File server starting
Thu Sep 10 17:45:11 2009 afs_krb_get_lrealm failed, using icequake.net.
Thu Sep 10 17:45:11 2009 /var/lib/openafs/local/sysid: doesn't exist
Thu Sep 10 17:45:11 2009 Creating new SysID file
Thu Sep 10 17:45:13 2009 VL_RegisterAddrs rpc failed; will retry periodic=
ally (code=3D5376, err=3D0)
Thu Sep 10 17:45:13 2009 Set thread id 14 for FSYNC_sync
Thu Sep 10 17:45:13 2009 FSYNC_sync: bind failed with (98), removed bogus=
/var/lib/openafs/local/fssync.sock
Thu Sep 10 17:45:13 2009 Partition /vicepa: attaching volumes
Thu Sep 10 17:45:14 2009 Partition /vicepa: attached 139 volumes; 0 volum=
es not attached
Thu Sep 10 17:45:14 2009 Set thread id 15 for 'FiveMinuteCheckLWP'
Thu Sep 10 17:45:14 2009 Set thread id 16 for 'HostCheckLWP'
Thu Sep 10 17:45:14 2009 Set thread id 17 for 'FsyncCheckLWP'
Thu Sep 10 17:45:14 2009 Getting FileServer name...
Thu Sep 10 17:45:14 2009 FileServer host name is 'valhalla'
Thu Sep 10 17:45:14 2009 Getting FileServer address...
Thu Sep 10 17:45:14 2009 FileServer valhalla has address 10.0.1.232 (0xe8=
01000a or 0xa0001e8 in host byte order)
Thu Sep 10 17:45:14 2009 File Server started Thu Sep 10 17:45:14 2009
This is on fileserver 2 (xanadu):
# cat /var/lib/openafs/local/NetInfo=20
10.0.1.230
f 65.38.17.159
# udebug valhalla 7002
Host's addresses are: 10.0.1.232 65.38.17.158=20
Host's 10.0.1.232 time is Thu Sep 10 17:55:48 2009
Local time is Thu Sep 10 17:55:48 2009 (time differential 0 secs)
Last yes vote for 230.1.0.10 was 16 secs ago (sync site);=20
Last vote started 13 secs ago (at Thu Sep 10 17:55:35 2009)
Local db version is 1251646288.3
I am not sync site
Lowest host 10.0.1.230 was set 16 secs ago
Sync host 10.0.1.230 was set 16 secs ago
Sync site's db version is 1251646288.3
0 locked pages, 0 of them for write
# vos listaddrs
xanadu.icequake.net
valhalla.icequake.net
# cat /var/log/openafs/PtLog
Thu Sep 10 17:45:20 2009 Using 10.0.1.230 as my primary address
# cat /var/log/openafs/FileLog
Thu Sep 10 17:45:20 2009 File server starting
Thu Sep 10 17:45:20 2009 afs_krb_get_lrealm failed, using icequake.net.
Thu Sep 10 17:45:21 2009 /var/lib/openafs/local/sysid: doesn't exist
Thu Sep 10 17:45:21 2009 Creating new SysID file
Thu Sep 10 17:45:23 2009 VL_RegisterAddrs rpc failed; will retry periodic=
ally (code=3D5376, err=3D0)
Thu Sep 10 17:45:23 2009 Set thread id 28 for FSYNC_sync
Thu Sep 10 17:45:23 2009 FSYNC_sync: bind failed with (98), removed bogus=
/var/lib/openafs/local/fssync.sock
Thu Sep 10 17:45:23 2009 Partition /vicepa: attaching volumes
Thu Sep 10 17:45:23 2009 Partition /vicepa: attached 47 volumes; 0 volume=
s not attached
Thu Sep 10 17:45:23 2009 Set thread id 29 for 'FiveMinuteCheckLWP'
Thu Sep 10 17:45:23 2009 Set thread id 30 for 'HostCheckLWP'
Thu Sep 10 17:45:23 2009 Set thread id 31 for 'FsyncCheckLWP'
Thu Sep 10 17:45:23 2009 Getting FileServer name...
Thu Sep 10 17:45:23 2009 FileServer host name is 'xanadu'
Thu Sep 10 17:45:23 2009 Getting FileServer address...
Thu Sep 10 17:45:23 2009 FileServer xanadu has address 10.0.1.230 (0xe601=
000a or 0xa0001e6 in host byte order)
Thu Sep 10 17:45:23 2009 File Server started Thu Sep 10 17:45:23 2009
Interestingly, the PtLog on the second server doesn't have the same note
about seeing two IP addresses for the primary server. I tried
restarting just the ptserver on that server but the resulting log was
the same as before.
--=20
Ryan C. Underwood, <nemesis@icequake.net>
--wRRV7LY7NUeQGEoC
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
iD8DBQFKqYUTIonHnh+67jkRAkdhAKClBSriwuGaB3/083YESwairlZwjgCgoWbQ
v9gg6SPWEDCIumoCkuS1W8A=
=KrdU
-----END PGP SIGNATURE-----
--wRRV7LY7NUeQGEoC--