<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0//EN" "http://www.w3.org/TR/REC-html40/strict.dtd"> <html><head><meta name="qrichtext" content="1" /><style type="text/css"> p, li { white-space: pre-wrap; } </style></head><body style=" font-family:'Sans Serif'; font-size:10pt; font-weight:400; font-style:normal;"> On Wednesday 11 February 2009 20:04:46 Alireza Nematollahi wrote: > Hello, > > I think I did it in order: > 1- nothing is mounted > > > mount > > rootfs on / type rootfs (rw) > proc on /proc type proc (rw) > sysfs on /sys type sysfs (rw) > devpts on /dev/pts type devpts (rw) > tmpfs on /dev/shm type tmpfs (rw) > rpc_pipefs on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw) > nfsd on /proc/fs/nfsd type nfsd (rw) > > 2- deactivating VG/LVs > > > vgchange -an disk2 > > 3- stop the storage > > > sdparm -C stop /dev/sdb > > 4- checking for LVs on other disk > > > lvscan > > Isn't it the right order? I'm new to LVM and that's why I may be > confused... > > I don't like doing "echo 1 > /sys/block/sdb/device/delete" since I can't > get the device restarted when I do sdparm -C start /dev/sdb... > > Thanks > > > > ----- Original Message ----- > From: "Bryn M. Reeves" <bmr@redhat.com> > To: "Alireza Nematollahi" <alirezan@sfu.ca> > Cc: "device-mapper development" <dm-devel@redhat.com> > Sent: Wednesday, February 11, 2009 10:50:45 AM GMT -08:00 US/Canada Pacific > Subject: Re: [dm-devel] Stopped disk causes LVM to go crazy > > Alireza Nematollahi wrote: > > hi. Thanks for your reply. Well, I had tried that before, but no luck. I > > tried it again, and still no luck: > > You have to do it *BEFORE* you disable/remove the storage. Doing it > afterwards will just throw all the errors you are seeing since the LVM > tools can no longer access the device. > > You have to do things in the correct order; first unmount any file > systems, then de-activate LVs/VGs, then stop or remove the storage > hardware. > > If you do this in the wrong order you'll end up with the problems > you're seeing. > > For now, you can clean up /dev/sdb by deleting it via sysfs: > > echo 1 > /sys/block/sdb/device/delete > > Which will stop other commands that scan the disks from tripping over > the dead device. You can also manually delete any remaining device > mapper devices that reference the removed disk (see similar thread > yesterday). Add /dev/sdb to the filter so that it doesn't get scanned... lvm.conf: filter = [ "r/sdb/", "a/.*/" ] Best regards, ./npf > > Regards, > Bryn. </body></html>