Purpose: This document will instruct an engineer on how to properly reboot a Call Manager node and perform the necessary validation afterward to ensure the system is fully operational.
- Open a CLI session to the CUCM Publisher
- Enter the command “utils service list” to gather the running services.
- NOTE: Copy and paste the running services into a notepad and save the notepad. We will need to compare the running services before and after the reboot to ensure all services have started.
- Enter the command “utils dbreplication runtimestate” to verify the database replication between the Call Manager nodes is in sync.
- Confirm the replication setup is (2) Setup Complete. If it is not (2) Setup Complete, do not proceed with the reboot and troubleshoot the database replication issue. Below is a screenshot of what we are looking for.
- Enter the command “utils service list” to gather the running services.
- Gather the registered phone count via RTMT
- Open RTMT and connect to the Publisher Call Manager node (RTMT is available for download via Call Manager Administration Page > Application > PlugIns)
- Navigate to the Voice/Video Tab in RTMT
- Select Device Summary
- Be sure to take note of the Registered Phone Devices for both Call Manager node, as well as the entire cluster (screenshots below)
- Run a continuous ping to the CUCM server
- Running a continuous ping will give us an idea of when the server shuts down and when the server is back up.
- Return to the CLI session to the Publisher
- Enter the command “utils system restart”
- Keep an eye on the continuous ping for when timeouts begin.
- NOTE: When pings begin timing out, this means the server is shut down. Wait about 5-10 minutes and pings should start responding again.
- Once pings begin to respond, wait another 5 minutes and then attempt to CLI into the Publisher.
- Verify all services have started
- Via the CLI session, run the “utils service list” command and verify the running services match the running service list that was taken before the reboot.
- NOTE: Services can sometimes take a while to start. After you have regained access to the CLI, it could take another 15 minutes for all services to start. Be patient.
- If a service does not start, we will have to manually start it by going to the Cisco Unified Servicability > Tools > Feature / Network Services
- Return to RTMT
- Verify the registered phone count is the same as before beginning the maintenance.
- If there are quite a few phones unregistered, go through CUCM and identify the unregistered phones.