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@xxxxxxxxxx>
To: "Alireza Nematollahi" <alirezan@xxxxxx>
Cc: "device-mapper development" <dm-devel@xxxxxxxxxx>
Sent: Wednesday, February 11, 2009 10:50:45 AM GMT -08:00 US/Canada Pacific
Subject: Re: 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).
Regards,
Bryn.
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@xxxxxxxxxx>
To: "Alireza Nematollahi" <alirezan@xxxxxx>
Cc: "device-mapper development" <dm-devel@xxxxxxxxxx>
Sent: Wednesday, February 11, 2009 10:50:45 AM GMT -08:00 US/Canada Pacific
Subject: Re: 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).
Regards,
Bryn.
-- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
