Restore via single command (A2HA)

[punitmundra]

Aha! Link: https://chef.aha.io/epics/SH-E-243

[punitmundra]

With Current restore, we have to stop all the instance of automate and chef-server manually. Post that we trigger the restore command.

The purpose of this task is to simply the restore steps. In the current scenario, If we want to restore the backup, we need to follow these steps.

• We need go to each Frontend node (automate + chef_server) and stop all the services.
• Run the restore command from any one of the automate-frontend node.

To make restore works we need to close all the DB connection, i.e. the reason we are stopping all the services. FE node are not aware of the other node details i.e. the reason we can not run the restore command alone (in A2HA). Note: In case of automate(ES and PG are on same machine) we don't stop all the service manually, it will taken care automatically.

One Alternative approach is that we can write the script. which we can execute from the bastion host. script first stop all the services on all the FE nodes and then it will trigger the restore command from one of the automate-FE node.

Acceptance Criteria

• Script will able to perform the restore operation from the bastion
• Command support with automate-cluster-ctl
• Need to add the documentation section for the command

Definition of Done

• Script should run from the bastion host.
• Script read the ssh user details and key from the a2ha.rb file.

• Restore step will take good amount of time (we should consider this step before implementation).

  Testing

  • Need to test the script from bastion (Ubuntu/CentOS/RHEL 7,8)
  • Need to check, how much time restore command will take from bastion node. In case of error, how user will get to know restore fails.

Note :- Post this Implementation, Customer may raise the question that, why Only Restore command support from bastion not the create backup.