Restarting Frontend, Dispatcher, & RabbitMQ logs on Domino version 4.x

On Domino 4.x

To restart the dispatcher you must log in to the central server, delete its pod. A new one will be created to replace it.

kubectl delete pod \ -n {PLATFORM_NAMESPACE} --selector app.kubernetes.io/instance=nucleus,app.kubernetes.io/component=dispatcher

where {PLATFORM_NAMESPACE} will be found with kubectl get ns and you look for the namespace that ends in something like domino-platform  and <pod name> is the name of the pod you want the logs for, i.e. nucleus-dispatcher-XXXX or nucleus-frontend-XXXX where XXXX is some random string of digits that uniquely identify the pod.

To restart the frontend you must log in to the central server, delete its pod. A new one will be created to replace it.

1. kubectl get po --all-namespaces|grep frontend
2. Identify FE pod names
3. kubectl delete <1st FE pod name>
4. kubectl get po --all-namespaces|grep frontend
5. You should see the pod delete, and disappear, followed by a new one spinning up immediately to replace it. When this pod is in a running state # with 2/2 containers running you can delete the other pod the same way.
6. kubectl delete <2nd FE pod name>
7. Make sure the 2nd pod recreates properly and test UI access to ensure everything is running as it should be.

If you suspect an intermittent or otherwise unknown connectivity issue between the three RabbitMQ cluster nodes, a rolling restart can be attempted:

1. kubectl rollout restart -n <platform namespace> sts/rabbitmq-ha-ord
2. If there is no rabbitmq-ha-ord statefulset, you may need to try
3. kubectl rollout restart -n <platform namespace> sts/rabbitmq-ha