Am 28.02.2012 18:00, schrieb Derrick Brashear: > On Tue, Feb 28, 2012 at 11:55 AM, Berthold Cogel <co...@uni-koeln.de> wrote: >> Hello! >> >> I'm trying to get openafs to work on Fedora 16. The rpms from rpmfusion >> did work with older kernel versions but failed totally with recent >> kernels. So I've build my own rpms. >> I took the source packages for 1.6.1-0.pre2 and build an source rpm with >> the makesrpm.pl. Build and install worked well with some minor fixes to >> the files section in the spec file. >> >> System is a Dell Latitude E6520 laptop. Because of this an becaus we're >> using DHCP with fixed IPs in our local net, network manager is involved. >> >> After activating the client service and rebooting the system I get now ... >> >> [root@charon/var/log]$ systemctl status openafs-client.service >> openafs-client.service - OpenAFS Client Service >> Loaded: loaded (/lib/systemd/system/openafs-client.service; enabled) >> Active: failed since Tue, 28 Feb 2012 08:07:54 +0100; 8h ago >> Process: 1448 ExecStart=/usr/vice/etc/afsd $AFSD_ARGS (code=exited, >> status=1/FAILURE) >> Process: 1441 ExecStartPre=/sbin/modprobe openafs (code=exited, >> status=0/SUCCESS) >> Process: 1436 ExecStartPre=/bin/chmod 0644 /usr/vice/etc/CellServDB >> (code=exited, status=0/SUCCESS) >> Process: 1418 ExecStartPre=/bin/sed -n w/usr/vice/etc/CellServDB >> /usr/vice/etc/CellServDB.local /usr/vice/etc/CellServDB.dist >> (code=exited, status=0/SUCCESS) >> CGroup: name=systemd:/system/openafs-client.service >> └ 1459 /usr/vice/etc/afsd -fakestat -afsdb -debug >> >> Although failed, there are several afs processes running which can't be >> stopped or killed. > > as always, afsd wants to be shut down with afsd -shutdown. it will > happily ignore your KILL signals because they're not helpful. >
I did try 'systemctl stop openafs-client.service'. This should have called the lines in openafs-client.service: ExecStop=/bin/umount /afs ExecStop=/usr/vice/etc/afsd -shutdown ExecStop=/sbin/rmmod openafs But this didn't work. The afsd seemed to be in a state where -shutdown didn't work. And each restart triggered another bunch of processes/threads which couldn't be removed. The interesting part is that this morning the LAN interface got up some milliseconds earlier during the boot. And so the network manager could do his job early enough for the afsd. And now stop and start with systemctl works as expected. Berthold Cogel _______________________________________________ OpenAFS-info mailing list OpenAFS-info@openafs.org https://lists.openafs.org/mailman/listinfo/openafs-info