A corrupted MDSM client cannot cause the array to disappear

The array must respond on OOB mgmt for MDSM to show anything.

Your symptoms indicate the array is not responding — not the MDSM client.

 

Rebooting DOES clear TCP/IP sessions

Since nothing changed, the controller is not stuck in TCP exhaustion — it is not booting, exactly as Dell cases describe for failed controllers.

 

Drives not showing mean controller never initialized them

This does not mean disk failure.

 

Try a Reseat or Fail Over the Controller Modules , most likely fix

Your screenshots show the RAID Controller Module as Unresponsive and the OOB IP unreachable — the exact symptom Dell associates with a hung or failed controller.

 

• Pull one controller (not both) for 10 seconds
• Reinsert
• Wait 3–5 minutes
• If still unresponsive, repeat with the second controller

This is a recovery action for an unresponsive MD32xx controller.

 

If the controller comes back - do data and management recover immediately.

If it does NOT boot (no LEDs except power):

Controller has failed and will need replacement.

 

Next Connect via SERIAL to Diagnose Boot Problems

When both GUI and network management fail, Dell’s procedure is to connect over serial to view the boot sequence.

This allows you (or Dell) to confirm:

• Controller firmware corruption
• Controller OS not booting
• Hardware failure
• NVSRAM issues

How:

• Use a null‑modem serial cable
• 115200‑8‑N‑1 settings
• Reseat the controller while connected and watch output

If you see boot text → controller is alive but stuck (recoverable).

If you see nothing → controller hardware failure (replacement required).

 

Controller Replacement most likely fix

If reseating and serial access confirm the controller does not boot, then:

A controller module has failed

Drives and data are usually intact

Replacing one controller typically restores full access

Dell confirms that a failed controller produces Unresponsive state in MDSM just like your screenshots.

You may need to contact support directly. They can:

• Verify failure over serial
• Provide/authorize replacement controller
• Guide safe insertion to avoid write‑cache issues
• Re-sync controllers