Troubleshooting Tip: FortiGate HA cluster shutdown issue with RAID-enabled devices

[edyrmishi]

Description	This article describes how, when a RAID-enabled FortiGate joins an existing cluster, the primary unit shuts down automatically with the message 'secondary and primary have different hdisk status, Shutdown the box!'
Scope	FortiGate.
Solution	A known issue has been observed when a FortiGate with RAID enabled joins an existing cluster. The primary FortiGate in the cluster shuts down automatically and displays the following message: secondary and primary have different hdisk status, Shutdown the box! This occurs regardless of HA uptime, priority, or serial number. When a FortiGate with RAID enabled joins, it takes precedence over other cluster members and forces them to shut down. Reproduction Scenario: A new HA is cluster deployed with the HA port disabled or disconnected. FGT-Primary is configured with a priority of 128, uptime greater than 5 minutes, and RAID disabled. FGT-Secondary is configured with a priority of 64, uptime of 2 minutes, and RAID enabled. HA port connected/enabled - this causes the primary FortiGate to shut down. A power cycle is required to bring the primary FortiGate back online. This issue is has been reported under ID 1098192 and a fix will be provided in FortiOS 7.6.3. For FortiOS versions prior to this release, it is essential to verify and configure RAID settings properly before adding a FortiGate device to the cluster to avoid unexpected shutdowns. The following commands are helpful in identifying Disk and RAID status: List disk devices and partitions: Display information about all of the disks: Check the Raid status: Raid enabled: Raid disabled: More information can be found in the FortiGate Administration Guide.