Step F: Rebuild the cluster - AWS Elemental Conductor File

Step F: Rebuild the cluster

  1. Restore the database on primary conductor file node.

  2. If you have a secondary conductor in the cluster, you need to restore the database on secondary conductor file node.

  3. If you have a secondary conductor in the cluster, you need to add the secondary conductor to the cluster.

    1. From the Linux prompt, log in to the secondary conductor node with the elemental user credentials.

    2. Run the following commands:

      [elemental@hostname ~]$ cd /opt/elemental_se
[elemental@hostname ~]$ sudo ./configure -s -c -t -n -z -xeula

    3. You will see the following prompts:

      [elemental@hostname ~]$ Configure this node as a secondary Elemental Conductor in an existing cluster? [N] 
Y
[elemental@hostname ~]$ What is the hostname or IP address of the cluster management node? 
<hostname of primary conductor>
...
[elemental@hostname ~]$ Are you sure you want to continue connecting (yes/no/[fingerprint])? 
yes
[elemental@hostname ~]$ elemental@<hostname> password: 
<hostname of primary conductor>
...
[elemental@hostname ~]$ Trust this public key? [Y] 
Y

  4. If HA mode is enabled before, you need to re-enable it again.

  5. Restore the database on all the worker nodes.

  6. Add all the worker nodes back to the cluster

    1. From the Linux prompt, log in to the worker node with the elemental user credentials.

    2. Run the following commands:

      [elemental@hostname ~]$ cd /opt/elemental_se
[elemental@hostname ~]$ sudo ./configure -s -c -t -n -z -xeula

    3. After being prompted to add the worker node to the Conductor File cluster, you are asked to trust the public key from the conductor node(s). You must accept this for the Conductor File node to control the worker node.

    4. After trusting the public key, continue through the configuration prompts as normal.

  7. Use the configure script to add worker nodes to the cluster.

    Important

    If this is your first time adding worker nodes to a cluster you must add worker nodes to the Conductor File cluster via the CLI. Using the Conductor File web interface to discover the worker node and add to the cluster causes the worker node to go into a failed state.

    1. On the worker node, enter the following command to change the directory: 

      cd /opt/elemental_se

    2. Enter the following command to start the configurations script on the worker node:

      sudo ./configure

    3. After being prompted to add the worker node to the Conductor File cluster, you are asked to trust the public key from the conductor node(s). You must accept this for the Conductor File node to control the worker node.

    4. After trusting the public key, continue through the configuration prompts as normal.