Converting a 7-Mode 8.2.1 3220 to a C-mode switchless system seemed to be going well until I needed to add the second node.
The A head is running fine with the basic configuration and when running cluster setup on the B head it failed with:
Node check ....Error: Member node nas30a failed to inspect joining node's manifest. RPC: Port mapper failure - RPC: Timed out Restarting Cluster Setup This node's storage failover partner is already a member of a cluster. Storage failover partners must be members of the same cluster. The cluster setup wizard will default to the cluster join dialog.
After much searching and messing around with storage failover and cluster failover (until it wouldn’t re-enable as the heads weren’t in an HA paired cluster) I suddenly went back to basics.
Is there connectivity between the two heads’ cluster interfaces?
At that time the A head had two cluster ports and interfaces:
netapp-test::> network port show -role cluster Auto-Negot Duplex Speed (Mbps) Node Port Role Link MTU Admin/Oper Admin/Oper Admin/Oper ------ ------ ------------ ---- ----- ----------- ---------- ------------ nas30a e1a cluster up 9000 true/true full/full auto/10000 e1b cluster up 9000 true/true full/full auto/10000 2 entries were displayed. netapp-test::> network interface show -role cluster Logical Status Network Current Current Is Vserver Interface Admin/Oper Address/Mask Node Port Home ----------- ---------- ---------- ------------------ ------------- ------- ---- nas30a clus1 up/up 169.254.210.85/16 nas30a e1a true clus2 up/up 169.254.210.86/16 nas30a e1b true
The B head only had one. So I configured the second port and interface as I thought that the A head might receive traffic on the e1a interface but return it on the e1b head as they’re all on the same subnet.
After that can they ping each other on a cluster IP? No!
Ok so if ping won’t work then I can see why the cluster join fails! I then deleted the clus2 interface on each head and tested ping again:
netapp-test::> network inter delete -node nas30a -lif clus2 nas30b::> network inter delete -lif clus2 nas30b::> ping -node nas30b -destination 169.254.210.85 (network ping) 169.254.210.85 is alive nas30b::>
The the cluster join then worked!
So something to remember: only configure a single cluster interface on a node when creating a cluster and when experiencing issues test basic connectivity!