Portal subsystem: restarting pods

Restart Portal subsystem pods in your API Connect deployment.

1. If you are restarting pods after updating the portal-ca, complete the following steps:
   1. Delete all of the portal pods with the following command:
      In the command, namespace is the namespace where the Portal subsystem is installed:

      • For a one replica deployment profile:

        kubectl delete pods portal-sitename-db-0 portal-sitename-www-0 -n namespace

      • For a three replica deployment profile:

        kubectl delete pods portal-sitename-db-0 portal-sitename-db-1 portal-sitename-db-2 portal-sitename-www-0 portal-sitename-www-1 portal-sitename-www-2 -n namespace

   2. Reload the NGINX configuration by running the following command:

      kubectl -n namespace exec -it portal-tunnel -- openresty -s reload

2. If you updated any other certificate, you can perform a rolling restart on the pods by completing the following steps:
   1. Determine the statefulset names by running the following command:

      kubectl -n namespace get statefulset | grep CR_name

      Where:

      • namespace is the namespace where the Portal subsystem is installed.
      • CR-name is the name of the CR used for installing the Portal subsystem.
   2. Restart pods by running the appropriate kubectl commands, which are shown in Table 1.

      Use the statefulset name that you obtained in step 2a.

      Table 1. Portal pods and the commands to restart them

      Pod name	Restart command
      portal-local_sitename-db	kubectl rollout restart statefulset/statefulset_name
      portal-local_sitename-www	kubectl rollout restart statefulset/statefulset_name
      portal-tunnel	kubectl rollout restart statefulset/statefulset_name
      portal-remote_sitename-db	kubectl rollout restart statefulset/statefulset_name
      portal-remote_sitename-www	kubectl rollout restart statefulset/statefulset_name
      portal-nginx Attention: Restart this pod last.	kubectl rollout restart statefulset/statefulset_name

3. Verify that all Portal pods are ready by running the following command:

   kubectl -n namespace get po 

   Where namespace is the namespace where the Portal subsystem is installed.

   The restart is complete when all pods are Running and Ready.