i have 5 nodes vcs cluster with CFS,parallel cluster.
i like to know the steps when the vxdg import -s DG is issued on master node:
-the node device tree is created at OS level
-the node device tree is created at vxvm level
-the disks,subdisks,plexes,volumes are created in RAM.
On the rest of the nodes i need to perform vxgd import -s and happens the same.
What should i do if an certain node i can not see a certain DG,this DG is seen on the master node.
If you are unable to see certain DGs on certain CFS cluster nodes, then you should check CFS cluster membership and make sure membership is fine. If cluster membership is fine, then check disks that are associated with DGs are visible from all the nodes from VxVM as well as OS level. In case OS cant see the disks then you will need to work on the OS front make it visible and then rescan disks from VxVM level.
if a DG can not be imported on a certain node and can be imported on the rest ofb the nodes,then could i try to import DG with vxdg -f import or vxdg -C import?
if i have fencing then i could try to delete registration key?
If you're unable to see a CFS diskgroup on a cluster node, check its cluster membership and rescan the disks. Do not manually import the disk group, it should only be done via cluster resources. Try adding the node to cluster membership again and modifying the diskgroup service group to start on that node.