[LU-2592] replay-vbr test 0c: error: set_param: /proc/{fs,sys}/{lnet,lustre}/*/lustre-MDT0000/sync_permission: Found no match

Mon, 18 Mar 2013 05:54:11 +0000

While running replay-vbr with failover configuration and hard failure mode (power off and on), test 0c failed as follows:

== replay-vbr test 0c: VBR: open (non O_CREAT) does not checks versions ============================== 11:36:50 (1357673810)
CMD: client-32vm3 /usr/sbin/lctl set_param *.lustre-MDT0000.sync_permission=0
client-32vm3: error: set_param: /proc/{fs,sys}/{lnet,lustre}/*/lustre-MDT0000/sync_permission: Found no match
CMD: client-32vm1.lab.whamcloud.com mkdir -p -m 755 /mnt/lustre/d0.replay-vbr/d0
CMD: client-32vm1.lab.whamcloud.com openfile -f O_RDWR:O_CREAT -m 0644 /mnt/lustre/d0.replay-vbr/d0/f0c
Succeed in opening file "/mnt/lustre/d0.replay-vbr/d0/f0c"(flags=O_RDWR, mode=644)
CMD: client-32vm3 sync
Filesystem           1K-blocks      Used Available Use% Mounted on
client-32vm3:client-32vm7:/lustre
                      36535940   1696868  32982672   5% /mnt/lustre
CMD: client-32vm3 /usr/sbin/lctl --device %lustre-MDT0000 notransno
client-32vm3: opening /dev/obd failed: No such device
client-32vm3: hint: the kernel modules may not be loaded
client-32vm3: No device found for name lustre-MDT0000: No such device
CMD: client-32vm3 /usr/sbin/lctl --device %lustre-MDT0000 readonly
client-32vm3: opening /dev/obd failed: No such device
client-32vm3: hint: the kernel modules may not be loaded
client-32vm3: No device found for name lustre-MDT0000: No such device
CMD: client-32vm3 /usr/sbin/lctl mark mds REPLAY BARRIER on lustre-MDT0000
client-32vm3: opening /dev/lnet failed: No such device
client-32vm3: hint: the kernel modules may not be loaded
client-32vm3: IOC_LIBCFS_MARK_DEBUG failed: No such device
CMD: client-32vm5 chmod 777 /mnt/lustre/d0.replay-vbr/d0
CMD: client-32vm5 chmod 666 /mnt/lustre/d0.replay-vbr/d0/f0c
CMD: client-32vm1.lab.whamcloud.com rm -f /tmp/multiop_bg.pid.16759 && MULTIOP_PID_FILE=/tmp/multiop_bg.pid.16759 LUSTRE= runmultiop_bg_pause /mnt/lustre/d0.replay-vbr/d0/f0c o_c
multiop /mnt/lustre/d0.replay-vbr/d0/f0c vo_c
TMPPIPE=/tmp/multiop_open_wait_pipe.20009
CMD: client-32vm1.lab.whamcloud.com cat /tmp/multiop_bg.pid.16759
node client-32vm1.lab.whamcloud.com multiop_bg started multiop_pid=20017
CMD: client-32vm5 grep -c /mnt/lustre' ' /proc/mounts
Stopping client client-32vm5 /mnt/lustre (opts:)
CMD: client-32vm5 lsof -t /mnt/lustre
CMD: client-32vm5 umount  /mnt/lustre 2>&1
Failing mds on node client-32vm3
CMD: client-32vm3 lctl dl
client-32vm3: error: dl: No such file or directory opening /proc/fs/lustre/devices
client-32vm3: opening /dev/obd failed: No such device
client-32vm3: hint: the kernel modules may not be loaded
client-32vm3: Error getting device list: No such device: check dmesg.
+ pm -h powerman --off client-32vm3
pm: warning: server version (2.3.5) != client (2.3.12)
Command completed successfully
affected facets:
+ pm -h powerman --on client-32vm3
pm: warning: server version (2.3.5) != client (2.3.12)
Command completed successfully
df pid is 20158
CMD: hostname 
pdsh@client-32vm1: gethostbyname("hostname") failed

The MDS failover pair is client-32vm3 and client-32vm7. And in the previous test 0b, MDS client-32vm3 has failed over to client-32vm7. So, in test 0c, the active MDS should be client-32vm7 instead of client-32vm3.

This is a test script issue under failover configuration.

Whamcloud Community JIRA

[LU-2592] replay-vbr test 0c: error: set_param: /proc/{fs,sys}/{lnet,lustre}/*/lustre-MDT0000/sync_permission: Found no match