IP Metrocluster recovery

[Northman]

We had just setup our new IP Metrocluster with FAS8230 when I accidentally did a "system power off" on one of the 4 nodes.  The node has now been taken over by its partner node on the same site. I looked through the options and it seems "Storage node giveback" is the way to go, but it did not work, because it did not have contact with the partner.

Any suggestions what to do before I contact support tomorrow? I am aware there is little info available, but you know what I did, so maybe you also know what to do.

[paul_stejskal]

Dumb question, but did you try to power on the system? You should be able to go into SP/BMC and do something like "system console" to confirm it is responding, or check status there. You can also check "system power status".

https://kb.netapp.com/Advice_and_Troubleshooting/Data_Storage_Systems/FAS_Systems/What_is_a_Baseboard_Management_Controller_(BMC)_and_how_do_I_use_it%...

[Northman]

yes, it is powered on. I am thinking maybe I have to do a "power cycle" or "system reset". The system reset will reset to last working state, it seems.

[Northman]

Here are more details:

MC-01::> storage failover failover show-giveback

Partner

Node Aggregate Giveback Status

-------------- ----------------- ---------------------------------------------

MC-01-01

CFO Aggregates Disk inventory not exchanged, automatic

giveback deferred

B21_aggr2 Not attempted yet

Warning: Unable to list entries on node MC-01-02. RPC: Couldn't make

connection [from mgwd on node "MC-01-01" (VSID: -1) to mgwd at

169.254.155.46]

2 entries were displayed.

MC-01::> storage failover show-giveback

Takeover

Node Partner Possible State Description

-------------- -------------- -------- -------------------------------------

MC-01-01 MC-01-02 false In takeover, Auto giveback deferred

MC-01-02 MC-01-01 - Unknown

2 entries were displayed.

Command to try (storage giveback failover giveback)

MC-01::> storage giveback failover giveback -ofnode MC-01-0

MC-01-01 MC-01-02

MC-01::> storage failover giveback -ofnode MC-01-02

[Northman · Accepted solution]

A power cycle fixed it.