Replacing a SPU in an HP host

Answer

1. Determine which SPU is failed by running the following command:

nzinventory | grep -i failed

The output would look similar to:
SPU 1192 Failed Paused (Now) 22 4 10.0.22.4 372.61 GB

nzspu locate -all -off
nzspu locate -id spuID 


3. Once the failed SPU is physically identified, remove it and insert the new one. There is no downtime required for this activity to be performed. No need for nzstop or nzystem pause etc. 

4. Once the new SPU is inserted and a short duration allowed for the SPU to boot up and come on-line, run the following command:

nzinventory | grep 10.0.22.4 

{Note: 10.0.22.4 is the I.P. Address of the SPU obtained from step 1 above}
The output of the above should return 2 SPUs with the same I.P. Address similar to:

SPU 1192 Failed Paused (Now) 22 4 10.0.22.4 372.61 GB
SPU 1305 Mismatched Initialized 22 4 10.0.22.4 372.61 GB

nzspu activate -id 1305 

    {Note: 1305 is the id of the newly inserted SPU obtained in Step 4 above}

6. Delete the failed SPU using:

nzspu delete -id 1192 

{Note: 1192 is the failed SPU. It should only be deleted after it has been physically removed from the system.}

7. Check to make sure the failed SPU is no longer present using:

nzinventory | grep -i failed

8. You can now turn off the LEDs on the SPUs using:

nzspu locate -all -off

9. Using a return Fedex shipping label provided in the spares kit, please send back the failed SPU to us for further analysis in-house.