SFHA 5.1 SP2 , Win2008 64-bit, with sql2008 and disaster recovery configuration with global cluster manager
The point is that due to unknown disk error during evacuating a cluster disk, a cluster disk group in the recovery site cannot be imported because of missing volumes.
Is there a way of solving this (eg. force import the disk group, and reconfigure vvr rvg group) , or do i have to reconfigure this cluster disk group (VVR and so on) from the beginning ?
Thanks in advanced
Is it just the volumes that are showing as missing, or do you have missing disks?
If you just see missing volumes, then usually all you have to do is right click the volume and select "Reactivate volume". This should correct your issue.
If you have missing disks, you may need to open a case with Technical Support.
Let us know if that works.
I agree, missing volumes most probably will happen because of missing disks. Have a look at error log & find if there missing disks.
Once your missing disks are back, you need to ensure OS can see the disks. Once OS is able to see the disks back, reactivating volume should bring back volumes to active state
Thanks all for the reply
The problem happened after one disk imported to the disk group, and tried to evacuate to this another working disk. There was four small volumes in it ,and after pressing enter for the evacuation, the source disk appeared to have only one volume, and the rest three ones were on a missing disk with a red exclamation mark.
After rescaning refreshing , presenting and unpsresenting the disk to the server, we end up in a disk group with missing volumes and probably missing disks (but the original are present) and unable to import it.
How can i force import this group, in order to reconfigure the RVG ?
Have in mind that this storage group is on DR server
Most likely due to presenting and unpresenting LUNs to the server, you have some missing disk records in the diskgroup. But this could have also happened prior to the evacuate. Be very careful when working with missing disks, especially if you are considering deleting those missing disk records.
In order to force import the diskgroup, use the following command wuthout the brackets: vxdg -g[DiskGroupName] -f import
Sometimes this will resolve your issue and the missing disks will disappear. If it does not, and you are having issues with the missing disk records, please consider open a Technical Support case so that we may investigate accordingly. If it is something simple, we may be able to assist you here.
[edited for spelling]
can you first clarify some points?
1)Output on both sides of
#vxdisk -oalldgs list
what exact commands did you run?
Evacuate should only remove the volume/subdisks/plexes etc after data has been copied successfully.
If I understand you correctly the volume is fine on the primary site, right?
In that case (and if all disks are visible) and especially if you moved a VVR object I think the best solution would be to destroy the RVG on the remote site and create from scratch and then do a full resync.
I found this in the SFU docs, but I think this should also be applicable for the SFW VVR:
Veritas Volume Replicator (VVR) objects cannot be moved between disk groups.
We are going to recreate the vvr configuration from scratch
Is there any safe way to do this?
I mean that we cannot stop the cluster at the primary site.
So how ca i destroy and reconfigure the RDS from the primary in order to do a full resync?