plex at TEMPRMSD state
hello,
one of the plexes I have under the volume is in TEMPRMSD stete.
The volume is holding data so I must keep the data.
The disks status is:
DEVICE TYPE DISK GROUP STATUS
disk_0 auto:ZFS - - ZFS
disk_1 auto:ZFS - - ZFS
disk_2 auto:ZFS - - ZFS
st2540-2_0 auto:cdsdisk - - online
st2540-2_1 auto:cdsdisk st2540-2_7 dg_backup online
st2540-2_2 auto:cdsdisk st2540-2_9 dg_data online
st2540-3_0 auto - - error
st2540-3_1 auto - - error
st2540-3_2 auto - - error
st2540-3_3 auto:cdsdisk st2540-3_3 dg_backup online
st2540-3_4 auto:cdsdisk st2540-3_4 dg_data online
- - st2540-3_4 dg_backup failed was:st2540-3_1
- - st2540-3_5 dg_data failed was:st2540-3_0
I initiated the following vxassist command and the prompt hangs.
vxassist -g dg_data mirror ora_vol st2540-3_4
the vxprint shows:
TY NAME ASSOC KSTATE LENGTH PLOFFS STATE TUTIL0 PUTIL0
dg dg_data dg_data - - - - - -
dm st2540-2_9 st2540-2_2 - 733904640 - - - -
dm st2540-3_4 st2540-3_4 - 733968128 - - - -
dm st2540-3_5 - - - - NODEVICE - -
v ora_vol fsgen ENABLED 733898752 - ACTIVE ATT1 -
pl ora_vol-01 ora_vol ENABLED 733898752 - ACTIVE - -
sd st2540-2_9-01 ora_vol-01 ENABLED 733898752 0 - - -
pl ora_vol-02 ora_vol DISABLED 733898752 - NODEVICE - -
sd st2540-3_5-01 ora_vol-02 DISABLED 733898752 0 NODEVICE - -
pl ora_vol-03 ora_vol ENABLED LOGONLY - ACTIVE - -
sd st2540-2_9-02 ora_vol-03 ENABLED 4160 LOG - - -
pl ora_vol-04 ora_vol ENABLED(SPARSE) LOGONLY - ACTIVE - -
sd st2540-3_5-02 ora_vol-04 DISABLED 4160 LOG NODEVICE - -
pl ora_vol-05 ora_vol ENABLED 733898752 - TEMPRMSD ATT -
sd st2540-3_4-01 ora_vol-05 ENABLED 733898752 0 - - -
Looking at the storage, I can see that the drives in the storage are blinking so I assume that the mirroring is in progress.
My questions:
1) Does the mirroring is really is process?
2) Is there a way to know what is the exact progress of the mirror process (percentage) or estimation when it will be over?
3) Will the state TEMPRMSD change to ACTIVE once the mirror is completed or should I take action to fix/resolve the TEMPRMSD state?
4) I found that in order to fix/resolve the TEMPRMSD state, the following should be executed:
vxmend -g dg_name clear all volume plex
vxplex -g dg_name det plex
vxplex -g dg_name att volume plex
If the above should be executed, please advise:
4a) The "clear all volume plex" is worriying me. Will I lose the data (will it wipe the existing good flex as well)?
4b) What will be the exact commands? Will it be:
vxmend -g dg_data clear all ora_vol plex ora_vol-05
vxplex -g dg_data det ora_vol-05
vxplex -g dg_data att ora_vol ora_vol-05
- Please run "vxtask list" command check the progress of the mirroring task.
- The TEMPRMSD state will be cleared when the task complets 100% . The mirror command will come out once task is finished.
- You don't need to run any "clear" command here.
- Even your another failed mirror will come back and re-attached once the storage comes back. I can see the pelx is detached due to storage failures.
- To void the vxassist command stuck (it is not hung, just waiting for mirroring to finish), please use "-b" option to push the actual mirroring operation in background.
