yes, I think that is probably a bug.
We may want to add to the man page an explanation that errors are
ignored if the option is used...
brassow
On Jan 22, 2010, at 9:56 AM, Mikulas Patocka wrote:
Hi
In RHEL 5.5, when executing "lvchange --monitor n -ay vg2/test_long",
dmevetd is not started, but lvm still passes "block_on_error" flag to
dm-raid1 target.
The table looks like this.
vg2-test_long: 0 65536 mirror disk 3 253:4 1024 block_on_error 2
253:5 0
253:6 0
I think it is a bug that can cause deadlocks, because if dmeventd
doesn't
handle the failure and "block_on_error" flag is used the I/Os are
blocked
forever.
Should I file a RHEL bug for this? Or do you have some other
explanation
for it?
LVM version: 2.02.56(1)-RHEL5 (2009-11-24)
Library version: 1.02.39-RHEL5 (2009-11-11)
Driver version: 4.11.5
Mikulas
--
dm-devel mailing list
dm-devel@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/dm-devel
