local disk restripe on storevirtual 12.5 nodes

[ssoul]

I have a 6 node 4500G2 cluster with (5 managers running) at 95%+ capacity with all LUNS thick provisioned and network RAID10.  The local raid on each box is currently RAID10, giving us 19TB usable.  

I'm looking to change the local RAID to RAID5 on each node (remove from cluster/management group, change raid to raid5, add back to management group/cluster, wait for restripe to finish, repeat) to get some additional capacity, but due to capacity issues I'm unsure if I can remove a node from the cluster/management group. 

I don't have a cluster with enough space available to test, but when a single node is removed from a cluster does the available space decrease when using network RAID10? I assume the luns affected by a single node leaving the cluster will just be in unprotected state until the node returns and restripes?

[Assen]

Hi ssoul,

on 4500g2/g1 nodes the raid setup should be as follows:

logical drive 0 - contains OS in RAID 10

logical drive 1 - contains DATA in RAID 5

logical drive 2 - contains DATA in RAID 5

it would be very strange if DATA LD are in RAID 10. Please check in CMC what the DATA RAID is.

You do not need to remove node from cluster to do reconfiguration. You could put the node in repair mode and after reconfig you 

should exchange the node back with the IP placeholder. 

BR

Assen