sanity test 160j tries to read the changelog after one of two
client mounts is unmounted. In this case, we can fail to read
the changelog and get a "Cannot send after transport endpoint
shutdown" error.
The intention of sanity test 160j is to check that
there is no LBUG due to missed obd device. So, do not try to
read from the changelog after file system unmount.
Test-Parameters: trivial testlist=sanity env=ONLY=160j,ONLY_REPEAT=200
Signed-off-by: James Nunez <jnunez@whamcloud.com>
Change-Id: I1746a422b25d546b9aae38ae8438d9c08bce8827
Reviewed-on: https://review.whamcloud.com/41317
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Mike Pershin <mpershin@whamcloud.com>
Reviewed-by: John L. Hammond <jhammond@whamcloud.com>
umount $MOUNT
stack_trap "mount_client $MOUNT" EXIT
- # read changelog
- cat <&4 >/dev/null || error "read changelog failed"
+ # read changelog, which may or may not fail, but should not crash
+ cat <&4 >/dev/null
# clear changelog
local cl_user="${CL_USERS[$SINGLEMDS]%% *}"
printf 'clear:'$cl_user':0' >&3
}
-run_test 160j "client can be umounted while its chanangelog is being used"
+run_test 160j "client can be umounted while its chanangelog is being used"
test_160k() {
[ $PARALLEL == "yes" ] && skip "skip parallel run"