Troubleshooting Node down and unresponsive via console or Service Processor

Node fails to boot and serial console connection may yield no output
- Node may or may not be taken over by partner
If SP/BMC is unresponsive over SSH
- connect to console port with serial cable and hit Enter to determine if console is responsive/at LOADER
- If there is no output, attempt to toggle to SP/BMC with CTRL+G from a serial console connection and collect the same log output as above

Note: To exit back to the console from the SP/BMC, use CTRL+D. For A700s systems, use Esc+T to return to the console instead

Connection via SP/BMC may also fail
- Terminal screen is blank
Physical inspection may show the controller has no lit LEDs and appears powered off despite connections to known good power
- Cable and port swaps do not impact the LED state

If SP/BMC access is available, attempt to connect and gather logs:
Powercycle of SP/BMC or node does not fix the issue
- How to use the SP or BMC to non-disruptively powercycle a node
Once logs have been collected or deemed impossible to collect, attempt a physical power cycle
- The BMC may be in a hung state. Perform a "cold boot" by completely removing power from the node (unplug both power cables) for at least 3 minutes, then re-apply power. This forces a hard reset of the BMC hardware.
- On some Netapp ONTAP system models, the SP/BMC is a seperate FRU that can be re-seated separate from the motherboard

Check for KCS/IPMI errors
- If the system is at the LOADER prompt or failing to boot, look for errors indicating "KCS cmd failed" or "BMC unresponsive"
- If these persist after a hard power cycle and firmware update, this strongly suggests a hardware defect on the management module/motherboard