when ZS3-4 cluster takeover/failover frequently within 1~2 days, what is checkpoint to fix problem ?

Hello Exports.

My customer has ZS3-4 ZFSSA (active/active).

This ZS3-4 was some problems.

1st. node1 4x4 SAS HBA  had a fault status. (PCIE-8 slot)

  Error message : The transmitting device sent an invalid request.

     node1 was takeovered to node2 at that time.

     -> Replaced it. and marked repaired manually.

2nd.  node2 4x4 SAS HBA had a fault status (PCIE-8 slot) /  the 1st node1's SAS HBA fault since two months later

  Error message : The transmitting device sent an invalid request.

     node2 was takeovered to node1 at that time.

     -> Replaced it. and marked repaired manually.

3rd. node1 was panic. and CPU0/P0 was fault status. ILOM detect its fault. But AKD cannot detected as fault.