Ciao, On Wed, Apr 15, 2009 at 01:37:40PM +0200, Cristina Bulfon wrote: > > On Apr 15, 2009, at 1:18 PM, Dejan Muhamedagic wrote: > >> Ciao, >> >> On Wed, Apr 15, 2009 at 12:53:41PM +0200, Cristina Bulfon wrote: >>> Ciao Dejan, >>> >>> I am doing back & forth on this item :-) >>> I moved to 2.14. version and back to V1 style... I don't use anymore >>> DRBD, >>> just the mount >> >> Do you need drbd? > > No.. when I started the first time to use heartbeat I couldn't manage the > filesystem mount with heartbeat > so I used DRDB as workaround, I don't need it since my devices are visible > through the SAN.
OK. Make sure that you also configure fencing/stonith! >>> So the haresources file is the follows >>> >>> afsitfs3.roma1.infn.it IPaddr::141.108.26.31/24/eth0 >>> afsitfs3.roma1.infn.it Filesystem::/dev/AFS/sda3::/vicepa::xfs >>> afsitfs3.roma1.infn.it Filesystem::/dev/AFS/sda1::/usr/afs::ext3 >>> afsitfs3.roma1.infn.it 141.108.26.31 afs >>> >>> when I put the master node in stand_by or I stop the heartbeat, happens >>> the >>> following things >>> >>> - try the umount the filesystems before to stop "afs".. >> >> Isn't it afs stop before filesystem? > > That's is the problem I don't understand why .. it seems that > the stop is performed in the same "start" order That can't be. Really. Can't recall anymore how v1 works, perhaps it looks at the status before deciding whether to stop a resource. >>> umount: /vicepa: device is busy >>> umount: /vicepa: device is busy >>> Filesystem[3427]: 2009/04/14_09:16:52 ERROR: Couldn't unmount >>> /vicepa; trying cleanup with SIGTERM >>> /vicepa: >> >> This may be normal, i.e. there could be processes using the >> filesystem, though typically there are only applications which >> depend on the filesystem (in this case afs) which should be >> doing something there. If this is a concern, you should check >> which processes have files open over there (fuser,lsof). >> >>> With 2.1.3 version I didn;t see any kind of those message, everything is >>> V1 >>> style was fine. >> >> I suspect that the afs RA is not working correctly, in particular >> the status operation. > I will take a look > > thanks cristina Thanks, Dejan _______________________________________________ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
