8.4. Replacing Controller Nodes

In certain circumstances a Controller node in a high availability cluster might fail. In these situations, you must remove the node from the cluster and replace it with a new Controller node. This also includes ensuring the node connects to the other nodes in the cluster.
This section provides instructions on how to replace a Controller node. The process involves running the openstack overcloud deploy command to update the Overcloud with a request to replace a controller node. Note that this process is not completely automatic; during the Overcloud stack update process, the openstack overcloud deploy command will at some point report a failure and halt the Overcloud stack update. At this point, the process requires some manual intervention. Then the openstack overcloud deploy process can continue.

Important

The following procedure only applies to high availability environments. Do not use this procedure if only using one Controller node.

8.4.1. Preliminary Checks

Before attempting to replace an Overcloud Controller node, it is important to check the current state of your Red Hat OpenStack Platform environment. Checking the current state can help avoid complications during the Controller replacement process. Use the following list of preliminary checks to determine if it is safe to perform a Controller node replacement. Run all commands for these checks on the Undercloud.
  1. Check the current status of the overcloud stack on the Undercloud:
    $ source stackrc
    $ heat stack-list --show-nested
    
    The overcloud stack and its subsequent child stacks should have either a CREATE_COMPLETE or UPDATE_COMPLETE.
  2. Perform a backup of the Undercloud databases:
    $ mkdir /home/stack/backup
    $ sudo mysqldump --all-databases --quick --single-transaction | gzip > /home/stack/backup/dump_db_undercloud.sql.gz
    $ sudo systemctl stop openstack-ironic-api.service openstack-ironic-conductor.service openstack-ironic-discoverd.service openstack-ironic-discoverd-dnsmasq.service
    $ sudo cp /var/lib/ironic-discoverd/inspector.sqlite /home/stack/backup
    $ sudo systemctl start openstack-ironic-api.service openstack-ironic-conductor.service openstack-ironic-discoverd.service openstack-ironic-discoverd-dnsmasq.service
    
  3. Check your Undercloud contains 10 GB free storage to accomodate for image caching and conversion when provisioning the new node.
  4. Check the status of Pacemaker on the running Controller nodes. For example, if 192.168.0.47 is the IP address of a running Controller node, use the following command to get the Pacemaker status:
    $ ssh heat-admin@192.168.0.47 'sudo pcs status'
    
    The output should show all services running on the existing nodes and stopped on the failed node.
  5. Check the following parameters on each node of the Overcloud's MariaDB cluster:
    • wsrep_local_state_comment: Synced
    • wsrep_cluster_size: 2
    Use the following command to check these parameters on each running Controller node (respectively using 192.168.0.47 and 192.168.0.46 for IP addresses):
    $ for i in 192.168.0.47 192.168.0.46 ; do echo "*** $i ***" ; ssh heat-admin@$i "sudo mysql --exec=\"SHOW STATUS LIKE 'wsrep_local_state_comment'\" ; sudo mysql --exec=\"SHOW STATUS LIKE 'wsrep_cluster_size'\""; done
    
  6. Check the RabbitMQ status. For example, if 192.168.0.47 is the IP address of a running Controller node, use the following command to get the status
    $ ssh heat-admin@192.168.0.47 "sudo rabbitmqctl cluster_status"
    
    The running_nodes key should only show the two available nodes and not the failed node.
  7. Disable fencing, if enabled. For example, if 192.168.0.47 is the IP address of a running Controller node, use the following command to disable fencing:
    $ ssh heat-admin@192.168.0.47 "sudo pcs property set stonith-enabled=false"
    
    Check the fencing status with the following command:
    $ ssh heat-admin@192.168.0.47 "sudo pcs property show stonith-enabled"
    
  8. Check the nova-compute service on the director node:
    $ sudo systemctl status openstack-nova-compute
    $ nova hypervisor-list
    
    The output should show all non-maintenance mode nodes as up.
  9. Make sure all Undercloud services are running:
    $ sudo systemctl -t service
    

8.4.2. Node Replacement

Identify the index of the node to remove. The node index is the suffix on the instance name from nova list output.
[stack@director ~]$ nova list
+--------------------------------------+------------------------+
| ID                                   | Name                   |
+--------------------------------------+------------------------+
| 861408be-4027-4f53-87a6-cd3cf206ba7a | overcloud-compute-0    |
| 0966e9ae-f553-447a-9929-c4232432f718 | overcloud-compute-1    |
| 9c08fa65-b38c-4b2e-bd47-33870bff06c7 | overcloud-compute-2    |
| a7f0f5e1-e7ce-4513-ad2b-81146bc8c5af | overcloud-controller-0 |
| cfefaf60-8311-4bc3-9416-6a824a40a9ae | overcloud-controller-1 |
| 97a055d4-aefd-481c-82b7-4a5f384036d2 | overcloud-controller-2 |
+--------------------------------------+------------------------+
In this example, the aim is to remove the overcloud-controller-1 node and replace it with overcloud-controller-3. First, set the node into maintenance mode so the director does not reprovision the failed node. Correlate the instance ID from nova list with the node ID from ironic node-list
[stack@director ~]$ ironic node-list
+--------------------------------------+------+--------------------------------------+
| UUID                                 | Name | Instance UUID                        |
+--------------------------------------+------+--------------------------------------+
| 36404147-7c8a-41e6-8c72-a6e90afc7584 | None | 7bee57cf-4a58-4eaf-b851-2a8bf6620e48 |
| 91eb9ac5-7d52-453c-a017-c0e3d823efd0 | None | None                                 |
| 75b25e9a-948d-424a-9b3b-f0ef70a6eacf | None | None                                 |
| 038727da-6a5c-425f-bd45-fda2f4bd145b | None | 763bfec2-9354-466a-ae65-2401c13e07e5 |
| dc2292e6-4056-46e0-8848-d6e96df1f55d | None | 2017b481-706f-44e1-852a-2ee857c303c4 |
| c7eadcea-e377-4392-9fc3-cf2b02b7ec29 | None | 5f73c7d7-4826-49a5-b6be-8bfd558f3b41 |
| da3a8d19-8a59-4e9d-923a-6a336fe10284 | None | cfefaf60-8311-4bc3-9416-6a824a40a9ae |
| 807cb6ce-6b94-4cd1-9969-5c47560c2eee | None | c07c13e6-a845-4791-9628-260110829c3a |
+--------------------------------------+------+--------------------------------------+
Set the node into maintenance mode:
[stack@director ~]$ ironic node-set-maintenance da3a8d19-8a59-4e9d-923a-6a336fe10284 true
Tag the new node as with the control profile.
[stack@director ~]$ ironic node-update 75b25e9a-948d-424a-9b3b-f0ef70a6eacf add properties/capabilities='profile:control,boot_option:local'
Create a YAML file (~/templates/remove-controller.yaml) that defines the node index to remove:
parameters:
  ControllerRemovalPolicies:
    [{'resource_list': ['1']}]

Important

If replacing the node with index 0, edit the heat templates and change the bootstrap node index and node validation index before starting replacement. Create a copy of the director's Heat template collection (see Chapter 10, Creating Custom Configuration and run the following command on the overcloud-without-mergepy.yaml file:
$ sudo sed -i "s/resource\.0/resource.1/g" ~/templates/my-overcloud/overcloud-without-mergepy.yaml
This changes the node index for the following resources:
ControllerBootstrapNodeConfig:
  type: OS::TripleO::BootstrapNode::SoftwareConfig
    properties:
      bootstrap_nodeid: {get_attr: [Controller, resource.0.hostname]}
      bootstrap_nodeid_ip: {get_attr: [Controller, resource.0.ip_address]}
And:
AllNodesValidationConfig:
  type: OS::TripleO::AllNodes::Validation
  properties:
    PingTestIps:
      list_join:
      - ' '
      - - {get_attr: [Controller, resource.0.external_ip_address]}
        - {get_attr: [Controller, resource.0.internal_api_ip_address]}
        - {get_attr: [Controller, resource.0.storage_ip_address]}
        - {get_attr: [Controller, resource.0.storage_mgmt_ip_address]}
        - {get_attr: [Controller, resource.0.tenant_ip_address]}
After identifying the node index, redeploy the Overcloud and include the remove-controller.yaml environment file:
[stack@director ~]$ openstack overcloud deploy --templates --control-scale 3 -e ~/templates/remove-controller.yaml [OTHER OPTIONS]

Important

If you passed any extra environment files or options when you created the Overcloud, pass them again here to avoid making undesired changes to the Overcloud.
However, note that the -e ~/templates/remove-controller.yaml is only required once in this instance.
The director removes the old node, creates a new one, and updates the Overcloud stack. You can check the status of the Overcloud stack with the following command:
[stack@director ~]$ heat stack-list --show-nested

8.4.3. Manual Intervention

During the ControllerNodesPostDeployment stage, the Overcloud stack update halts with an UPDATE_FAILED error at ControllerLoadBalancerDeployment_Step1. This is because some Puppet modules do not support nodes replacement. This point in the process requires some manual intervention. Follow these configuration steps:
  1. Get a list of IP addresses for the Controller nodes. For example:
    [stack@director ~]$ nova list
    ... +------------------------+ ... +-------------------------+
    ... | Name                   | ... | Networks                |
    ... +------------------------+ ... +-------------------------+
    ... | overcloud-compute-0    | ... | ctlplane=192.168.0.44   |
    ... | overcloud-controller-0 | ... | ctlplane=192.168.0.47   |
    ... | overcloud-controller-2 | ... | ctlplane=192.168.0.46   |
    ... | overcloud-controller-3 | ... | ctlplane=192.168.0.48   |
    ... +------------------------+ ... +-------------------------+
    
  2. Check the nodeid value of the removed node in the /etc/corosync/corosync.conf file on an existing node. For example, the existing node is overcloud-controller-0 at 192.168.0.47:
    [stack@director ~]$ ssh heat-admin@192.168.0.47 "sudo cat /etc/corosync/corosync.conf"
    
    This displays a nodelist that contains the ID for the removed node (overcloud-controller-1):
    nodelist {
      node {
        ring0_addr: overcloud-controller-0
        nodeid: 1
      }
      node {
        ring0_addr: overcloud-controller-1
        nodeid: 2
      }
      node {
        ring0_addr: overcloud-controller-2
        nodeid: 3
      }
    }
    
    Note the nodeid value of the removed node for later. In this example, it is 2.
  3. Delete the failed node from the Corosync configuration on each node and restart Corosync. For this example, log into overcloud-controller-0 and overcloud-controller-2 and run the following commands:
    [stack@director] ssh heat-admin@192.168.201.47 "sudo pcs cluster localnode remove overcloud-controller-1"
    [stack@director] ssh heat-admin@192.168.201.47 "sudo pcs cluster reload corosync"
    [stack@director] ssh heat-admin@192.168.201.46 "sudo pcs cluster localnode remove overcloud-controller-1"
    [stack@director] ssh heat-admin@192.168.201.46 "sudo pcs cluster reload corosync"
    
  4. Log into one of the remaining nodes and delete the node from the cluster with the crm_node command:
    [stack@director] ssh heat-admin@192.168.201.47
    [heat-admin@overcloud-controller-0 ~]$ sudo crm_node -R overcloud-controller-1 --force
    
    Stay logged into this node.
  5. Delete the failed node from the RabbitMQ cluster:
    [heat-admin@overcloud-controller-0 ~]$ sudo rabbitmqctl forget_cluster_node rabbit@overcloud-controller-1
    
  6. Delete the failed node from MongoDB. First, find the IP address for the node's Interal API connection.
    [heat-admin@overcloud-controller-0 ~]$ sudo netstat -tulnp | grep 27017
    tcp        0      0 192.168.0.47:27017    0.0.0.0:*               LISTEN      13415/mongod
    
    Check that the node is the primary replica set:
    [root@overcloud-controller-0 ~]# echo "db.isMaster()" | mongo --host 192.168.0.47:27017
    MongoDB shell version: 2.6.11
    connecting to: 192.168.0.47:27017/echo
    {
      "setName" : "tripleo",
      "setVersion" : 1,
      "ismaster" : true,
      "secondary" : false,
      "hosts" : [
        "192.168.0.47:27017",
        "192.168.0.46:27017",
        "192.168.0.45:27017"
      ],
      "primary" : "192.168.0.47:27017",
      "me" : "192.168.0.47:27017",
      "electionId" : ObjectId("575919933ea8637676159d28"),
      "maxBsonObjectSize" : 16777216,
      "maxMessageSizeBytes" : 48000000,
      "maxWriteBatchSize" : 1000,
      "localTime" : ISODate("2016-06-09T09:02:43.340Z"),
      "maxWireVersion" : 2,
      "minWireVersion" : 0,
      "ok" : 1
    }
    bye
    
    This should indicate if the current node is the primary. If not, use the IP address of the node indicated in the primary key.
    Connect to MongoDB on the primary node:
    [heat-admin@overcloud-controller-0 ~]$ mongo --host 192.168.0.47
    MongoDB shell version: 2.6.9
    connecting to: 192.168.0.47:27017/test
    Welcome to the MongoDB shell.
    For interactive help, type "help".
    For more comprehensive documentation, see
    http://docs.mongodb.org/
    Questions? Try the support group
    http://groups.google.com/group/mongodb-user
    tripleo:PRIMARY>
    
    Check the status of the MongoDB cluster:
    tripleo:PRIMARY> rs.status()
    
    Identify the node using the _id key and remove the failed node using the name key. In this case, we remove Node 1, which has 192.168.0.45:27017 for name:
    tripleo:PRIMARY> rs.remove('192.168.0.45:27017')
    

    Important

    You must run the command against the PRIMARY replica set. If you see the following message:
    "replSetReconfig command must be sent to the current replica set primary."
    
    Relog into MongoDB on the node designated as PRIMARY.

    Note

    The following output is normal when removing the failed node's replica set:
    2016-05-07T03:57:19.541+0000 DBClientCursor::init call() failed
    2016-05-07T03:57:19.543+0000 Error: error doing query: failed at src/mongo/shell/query.js:81
    2016-05-07T03:57:19.545+0000 trying reconnect to 192.168.0.47:27017 (192.168.0.47) failed
    2016-05-07T03:57:19.547+0000 reconnect 192.168.0.47:27017 (192.168.0.47) ok
    
    Exit MongoDB:
    tripleo:PRIMARY> exit
    
  7. Update list of nodes in the Galera cluster:
    [heat-admin@overcloud-controller-0 ~]$ sudo pcs resource update galera wsrep_cluster_address=gcomm://overcloud-controller-0,overcloud-controller-3,overcloud-controller-2
    
  8. Add the new node to the cluster:
    [heat-admin@overcloud-controller-0 ~]$ sudo pcs cluster node add overcloud-controller-3
    
  9. Check the /etc/corosync/corosync.conf file on each node. If the nodeid of the new node is the same as the removed node, update the value to a new nodeid value. For example, the /etc/corosync/corosync.conf file contains an entry for the new node (overcloud-controller-3):
    nodelist {
      node {
        ring0_addr: overcloud-controller-0
        nodeid: 1
      }
      node {
        ring0_addr: overcloud-controller-2
        nodeid: 3
      }
      node {
        ring0_addr: overcloud-controller-3
        nodeid: 2
      }
    }
    
    Note that in this example, the new node uses the same nodeid of the removed node. Update this value to a unused node ID value. For example:
    node {
      ring0_addr: overcloud-controller-3
      nodeid: 4
    }
    
    Update this nodeid value on each Controller node's /etc/corosync/corosync.conf file, including the new node.
  10. Restart the Corosync service on the existing nodes only. For example, on overcloud-controller-0:
    [heat-admin@overcloud-controller-0 ~]$ sudo pcs cluster reload corosync
    
    And on overcloud-controller-2:
    [heat-admin@overcloud-controller-2 ~]$ sudo pcs cluster reload corosync
    
    Do not run this command on the new node.
  11. Start the new Controller node:
    [heat-admin@overcloud-controller-0 ~]$ sudo pcs cluster start overcloud-controller-3
    
  12. Enable the keystone service on the new node. Copy the /etc/keystone directory from a remaining node to the director host:
    [heat-admin@overcloud-controller-0 ~]$ sudo -i
    [root@overcloud-controller-0 ~]$ scp -r /etc/keystone stack@192.168.0.1:~/.
    
    Log in to the new Controller node. Remove the /etc/keystone directory from the new Controller node and copy the keystone files from the director host:
    [heat-admin@overcloud-controller-3 ~]$ sudo -i
    [root@overcloud-controller-3 ~]$ rm -rf /etc/keystone
    [root@overcloud-controller-3 ~]$ scp -r stack@192.168.0.1:~/keystone /etc/.
    [root@overcloud-controller-3 ~]$ chown -R keystone: /etc/keystone
    [root@overcloud-controller-3 ~]$ chown root /etc/keystone/logging.conf /etc/keystone/default_catalog.templates
    
    Edit /etc/keystone/keystone.conf and set the admin_bind_host and public_bind_host parameters to new Controller node's IP address. To find these IP addresses, use the ip addr command and look for the IP address within the following networks:
    • admin_bind_host - Provisioning network
    • public_bind_host - Internal API network

    Note

    These networks might differ if you deployed the Overcloud using a custom ServiceNetMap parameter.
    For example, if the Provisioning network uses the 192.168.0.0/24 subnet and the Internal API uses the 172.17.0.0/24 subnet, use the following commands to find the node’s IP addresses on those networks:
    [root@overcloud-controller-3 ~]$ ip addr | grep "192\.168\.0\..*/24"
    [root@overcloud-controller-3 ~]$ ip addr | grep "172\.17\.0\..*/24"
    
  13. Enable and restart some services through Pacemaker. The cluster is currently in maintenance mode and you will need to temporarily disable it to enable the service. For example:
    [heat-admin@overcloud-controller-3 ~]$ sudo pcs property set maintenance-mode=false --wait
    
  14. Wait until the Galera service starts on all nodes.
    [heat-admin@overcloud-controller-3 ~]$ sudo pcs status | grep galera -A1
    Master/Slave Set: galera-master [galera]
    Masters: [ overcloud-controller-0 overcloud-controller-2 overcloud-controller-3 ]
    
    If need be, perform a `cleanup` on the new node:
              [heat-admin@overcloud-controller-3 ~]$ sudo pcs resource cleanup galera overcloud-controller-3
    
  15. Wait until the Keystone service starts on all nodes.
    [heat-admin@overcloud-controller-3 ~]$ sudo pcs status | grep keystone -A1
    Clone Set: openstack-keystone-clone [openstack-keystone]
    Started: [ overcloud-controller-0 overcloud-controller-2 overcloud-controller-3 ]
    
    If need be, perform a `cleanup` on the new node:
              [heat-admin@overcloud-controller-3 ~]$ sudo pcs resource cleanup openstack-keystone-clone overcloud-controller-3
    
  16. Switch the cluster back into maintenance mode:
    [heat-admin@overcloud-controller-3 ~]$ sudo pcs property set maintenance-mode=true --wait
    
The manual configuration is complete. Re-run the Overcloud deployment command to continue the stack update:
[stack@director ~]$ openstack overcloud deploy --templates --control-scale 3 [OTHER OPTIONS]

Important

If you passed any extra environment files or options when you created the Overcloud, pass them again here to avoid making undesired changes to the Overcloud.
However, note that the remove-controller.yaml file is no longer needed.

8.4.4. Finalizing Overcloud Services

After the Overcloud stack update completes, some final configuration is required. Log in to one of the Controller nodes and refresh any stopped services in Pacemaker:
[heat-admin@overcloud-controller-0 ~]$ for i in `sudo pcs status|grep -B2 Stop |grep -v "Stop\|Start"|awk -F"[" '/\[/ {print substr($NF,0,length($NF)-1)}'`; do echo $i; sudo pcs resource cleanup $i; done
Perform a final status check to make sure services are running correctly:
[heat-admin@overcloud-controller-0 ~]$ sudo pcs status

Note

If any services have failed, use the pcs resource cleanup command to restart them after resolving them.
Enable fencing if you disabled it during the node replacement. For example, if 192.168.0.47 is the IP address of a running Controller node, use the following command to enable fencing:
[heat-admin@overcloud-controller-0 ~]$ sudo pcs property set stonith-enabled=true
Exit to the director
[heat-admin@overcloud-controller-0 ~]$ exit

8.4.5. Finalizing Overcloud Network Agents

Source the overcloudrc file so that you can interact with the Overcloud. Check your routers to make sure the L3 agents are properly hosting the routers in your Overcloud environment. In this example, we use a router with the name r1:
[stack@director ~]$ source ~/overcloudrc
[stack@director ~]$ neutron l3-agent-list-hosting-router r1
This list might still show the old node instead of the new node. To replace it, list the L3 network agents in your environment:
[stack@director ~]$ neutron agent-list | grep "neutron-l3-agent"
Identify the UUID for the agents on the new node and the old node. Add the router to the agent on the new node and remove the router from old node. For example:
[stack@director ~]$ neutron l3-agent-router-add fd6b3d6e-7d8c-4e1a-831a-4ec1c9ebb965 r1
[stack@director ~]$ neutron l3-agent-router-remove b40020af-c6dd-4f7a-b426-eba7bac9dbc2 r1
Perform a final check on the router and make all are active:
[stack@director ~]$ neutron l3-agent-list-hosting-router r1
Delete the existing Neutron agents that point to old Controller node. For example:
[stack@director ~]$ neutron agent-list -F id -F host | grep overcloud-controller-1
| ddae8e46-3e8e-4a1b-a8b3-c87f13c294eb | overcloud-controller-1.localdomain |
[stack@director ~]$ neutron agent-delete ddae8e46-3e8e-4a1b-a8b3-c87f13c294eb

8.4.6. Finalizing Compute Services

Compute services for the removed node still exist in the Overcloud and require removal. Source the overcloudrc file so that you can interact with the Overcloud. Check the compute services for the removed node:
[stack@director ~]$ source ~/overcloudrc
[stack@director ~]$ nova service-list | grep "overcloud-controller-1.localdomain"
Remove the compute services for the node. For example, if the nova-scheduler service for overcloud-controller-1.localdomain has an ID of 5, run the following command:
[stack@director ~]$ nova service-delete 5
Perform this task for each service of the removed node.
Check the openstack-nova-consoleauth service on the new node.
[stack@director ~]$ nova service-list | grep consoleauth
If the service is not running, log into a Controller node and restart the service:
[stack@director] ssh heat-admin@192.168.201.47
[heat-admin@overcloud-controller-0 ~]$ pcs resource restart openstack-nova-consoleauth

8.4.7. Conclusion

The failed Controller node and its related services are now replaced with a new node.