9.4.2. Restarting the Red Hat Ceph Storage 5 Alertmanager service

Procedure

1. Log in to the OpenStack Controller node.

   참고

   Confirm that you are logged into a node that is hosting the Ceph Manager service. In default deployments, this is the OpenStack Controller node. You must be logged into a Controller node that is running the Ceph Manager service.

2. View the current Alertmanager specification file:

   $ sudo cephadm shell -- ceph orch ls --export alertmanager

3. Create a specification file for the Alertmanager service based on the output from the previous step.

   The following is an example of a specification file:

   service_type: alertmanager
   service_name: alertmanager
   placement:
     count: 3
     label: monitoring
   networks:
   - 10.10.10.0/24
   - 10.10.11.0/24

   참고

   The IP addresses in the networks list should correspond with the Storage/Ceph public networks in your environment.

4. Save the specification file as /root/alertmanager.spec.

5. Stop the HAProxy service:

   # pcs resource disable haproxy-bundle

6. Stop the Alertmanager service:

   # cephadm shell -k /etc/ceph/<stack>.client.admin.keyring -- ceph orch rm alertmanager

   • Replace <stack> with the name of your stack.

7. Start the Alertmanager service:

   # cephadm shell -k /etc/ceph/<stack>.client.admin.keyring -m /root/alertmanager.spec -- ceph orch apply -i /mnt/alertmanager.spec

   • Replace <stack> with the name of your stack.

8. Start the HAProxy service:

   # pcs resource enable haproxy-bundle