Check multipath.conf.annotated in the source. This will retry n times
before failing.
no_path_retry {n}
I have repeatable issues with setting this higher than about 3-5
however, where IO queues (more like blocks) forever, even after failing
the path back... requiring a reboot, and resulting in data corruption. YMMV.
/eli
Peter Koutoupis wrote:
I cannot seem to find an answer for this no matter where I look, but if
I wanted to queue all IO if there was no path for a LIMITED time before
it eventually times out, does a feature exist? Currently I am using "1
queue_if_no_path" but the last thing I want is for a path to be
unavailable and have it consume all host resources until the host cannot
handle it anymore. Thanks in advance.
Regards,
Peter Koutoupis
Systems Engineer
Xyratex International, Inc.
1804 Centre Point Circle
Suite 112
Naperville, IL 60563 USA
(630)364-7656
--
dm-devel mailing list
dm-devel@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/dm-devel
--
dm-devel mailing list
dm-devel@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/dm-devel
