Automated, Online: Restoring the Cluster

Before You Begin

• Make sure you have followed the Prerequisites.
• Make sure the backup is disabled on the backup cluster. Inconsistencies will occur if a new backup is created while restoring the cluster. For details, see Automated: Disabling the backup on the cluster.
• Make sure package wget, unzip, jq are available on all the restore nodes.
• All external data sources must be the same (SQL Server).
• You must restart the NFS server before cluster restore. To do that, run the following command on the NFS Server node: systemctl restart nfs-server.
• The restore cluster must have the same FQDN as the backup cluster.
• Make sure you have prepared your environment for using the uipathctl.sh script. For details, see Using uipathctl.sh.
• At the time of cluster restore, you must provide the inventory.ini file containing the IP address of the Ansible host node and information on how to SSH to all nodes. The uipathctl.sh script cannot generate this file automatically as the cluster is not present at the time of the restore process.

  Installation type	Configuration instructions	Requirements
  Online single-node evaluation mode	Configuring the machines Configuring the DNS	Download the following file: sf-installer.zip – Mandatory. See sf-installer.zip for download instructions.
  Online multi-node HA-ready production mode	Configuring the machines Configuring the load balancer Configuring the DNS	Download the following file: sf-installer.zip – Mandatory. See sf-installer.zip for download instructions.

Ansible automates the process of restoring the Automation Suite cluster. In the following illustration, the node Server 1, which is the Ansible host node, uses Ansible to orchestrate the restore of all the nodes in the cluster.

Preparing the Restore Process

Important: If you have multiple nodes in the cluster, follow these steps on all the nodes.

1. Log into the server node that you identified as the Ansible host node.
2. Create the /opt/UiPathAutomationSuite/{version}/installer folder to host the installer.

   # Become root
   sudo su -
   
   # Replace {version} with actual installer version
   
   mkdir -p /opt/UiPathAutomationSuite/{version}/installer
   chmod -R 777 /opt/UiPathAutomationSuite

3. Copy the restore.json file containing the restore configuration and other installer files to the /opt/UiPathAutomationSuite/{version}/installer folder. For details, see Preparing the restore configuration.

   # Replace {version} with actual installer version
   
   cp /path/to/restore.json /opt/UiPathAutomationSuite/{version}/installer
   cp /path/to/sf-installer.zip /opt/UiPathAutomationSuite/{version}/installer

4. Unzip the sf-installer.zip file.

   # Navigate to installer folder. Replace {version} with actual installer version
   cd /opt/UiPathAutomationSuite/{version}/installer
   
   unzip sf-installer.zip -d . 
   cd sf-installer

Restoring the Cluster

To restore the cluster, you must provide the following files to the uipathctl.sh script:

• inventory.ini – This file contains information on all the nodes in the cluster and SSH details. The uipathctl.sh script does not generate the inventory.ini file automatically as the cluster is not present at the time of cluster restore.
• restore.json – This file contains the restore configuration. For details, see Preparing the restore configuration section.

To restore the cluster, run the following command with the basic parameters. Make sure to use the parameter values applicable to you.

./uipathctl.sh restore --install-type online --restore-config ./restore.json --inventory ./inventory.ini

Basic Parameters

Parameter	Description
--install-type	Possible values: online and offline. Since this page provides restore instructions for online, choose the online value.
--restore-config	The file contains the restore configuration. For details, see Preparing the restore configuration.
--inventory	The file contains information on all the nodes in the cluster, and SSH details. For details, see Generating the Ansible inventory.ini file.

Advanced Parameters

Parameter

Description

--ansible-username

Specify the username that you will use for SSH connections to all the nodes. Defaults to the current user. If you use a different username for all the nodes, instead of using this parameter, set the username for all the nodes in inventory.ini and pass it to the script. Example: While running the script you are logged in as the myadminuser username. However, if you want to use the testadmin username to connect via SSH, you must provide testadmin to a value for this parameter.

Restoring Cluster_config.json

After cluster recovery, you may want to recover the cluster_config.json for future reference. You may need the file when adding new nodes to cluster, when upgrading, etc.

To restore cluster_config.json, run the following command from any of the server nodes:

./configureUiPathAS.sh config get -o ./cluster_config.json

Important:

The backup is not enabled on the restored cluster. If you wish to enable it, refer to Enabling the backup on the cluster.

Restoring the cluster is not an idempotent operation. In case of a failed restore, take one of the following steps before retrying the operation:

• Uninstall Automation Suite on all the nodes. For instructions, see Troubleshooting.
• Follow the manual restore steps. For instructions, see Manual, online: Restoring the cluster.