An NSX Edge node is unexpectedly found in Maintenance Mode, potentially impacting traffic or preventing configuration changes. The customer or administrator may be unaware of when or why the state change occurred.
General symptoms include:
Edge status reported as Maintenance Mode in the NSX Manager UI.
Routing services or Load Balancing potentially redirected to the peer node (in Active/Standby setups).
Discrepancies between external monitoring alerts and the actual state of the Edge node.
VMware NSX
Maintenance Mode on an NSX Edge is not a random event; it is explicitly triggered via the NSX Manager UI, a linked orchestration tool (like SDDC Manager or vRealize Automation), or a direct API call.
To identify the exact trigger and timestamp, review the logs on the affected Edge node and the NSX Manager.
Review Edge Node Logs to see when the edge node was first reported to be in Maintenance mode:
Log in to the Edge node CLI or review the log bundles. Look for the opsagent notifying the framework of the state change in /var/log/syslog :xx-xx-1xTxx:xx:xx.XXZ edge.xxx.xxx.xxx.net NSX 2642 - [nsx@6876 comp="nsx-edge" subcomp="opsagent" s2comp="edge-service" tid="2747" level="INFO"] GET /api/v1/node/configuration response {#012 "maintenance_mode_enabled": "enabled",#012
Identify the API Trigger
On the edge node, check to see if an API was fired to initiate edge node maintenance mode.
Search /var/log/nvpapi/api_access.log for the maintenance-mode string:
2XX6-XX-XXTX7:XX:XX.XXXZ INFO - 'PUT /api/v1/node/maintenance-mode --- New value: {"maintenance_mode": true, "notify_mpa": true}' 200 231 "" "" ""
If the Edge node remains stuck in Maintenance Mode or if the trigger cannot be identified and you need to restore service:
Attempt to exit Maintenance Mode via the NSX Manager UI: System > Fabric > Nodes > Edge Transport Nodes > [Select Node] > Actions > Exit Maintenance Mode.
If you are contacting Broadcom support about this issue, please provide the following:
NSX Manager support bundles.
NSX Edge support bundles from the affected node.
Specific timeframe (Date/Time/Timezone) when the node was first noticed to be in Maintenance Mode.
Handling Log Bundles for offline review with Broadcom support: