Configuring A Failover Server

The ThinkAutomation Professional Edition can be configured to act as a failover server for another ThinkAutomation instance. If the main ThinkAutomation Server goes offline, the failover server takes over message processing until the main server is back online.

Preparation

  • Your main ThinkAutomation server should be installed, configured and running.
  • You must use SQL Server, MySQL, PostgreSQL or MongoDB as your message store database.
  • The database must be available to the computer that will run the failover instance of ThinkAutomation. The database server would normally be on a separate computer to the ThinkAutomation main server.
  • The Backup Solution Settings In Message Store option must be enabled in the main ThinkAutomation server settings.
  • On the main ThinkAutomation server ensure the failover server IP address is in the Whitelist (Server Settings - Clients - Whitelist IP Addresses).
  • On the main ThinkAutomation server ensure port 9899 is open for inbound connections (from the failover server) on any firewalls.

Setting Up

On a separate computer, run the ThinkAutomation.exe setup again to install ThinkAutomation. After setup is complete the ThinkAutomation Node Configuration application will start.

Select the Failover Server option and click Next. You will then need to register this instance. Enter your registration details and serial number and click Next. You cannot use the same serial number that has been used to register the main ThinkAutomation instance.

You must then enter the computer name or IP address of the main ThinkAutomation computer. Click the Verify button to verify that the computer can connect to the ThinkAutomation main service.

Enable the Automatic Failover option if you want this instance to automatically monitor the main ThinkAutomation Instance.

How It Works

The failover ThinkAutomation server connects to the main ThinkAutomation server every few seconds to verify that it is running. If the main server cannot be reached (after a couple of retries) the failover server restores all current solution and global settings from the message store database backup and starts processing. Once the main server comes back online it stops processing and reverts to its waiting state.

During failover you can use the Studio to connect to the failover server to view processing status and to view the message store - however you cannot make changes to any settings.

Manual Failover

If the Automatic Failover option is not enabled then the failover instance of ThinkAutomation will not start processing if the main server cannot be contacted. You must enable the failover server manually. To do this, start the ThinkAutomation Studio on the failover server. After logging in you will be asked if you want to enable the failover. The failover server will still stop processing if the main server comes online even if Automatic Failover is disabled.

Note

If you use any Counter or Embedded Data actions in your automations then you must configure a shared location for the embedded database files. This is done using the EmbeddedStorePath registry key. This should be set to a path that is accessible by both the main ThinkAutomation Server and the Failover Server.

Tip

Parker Software can assist you in setting up a fault-tolerant, load-balanced ThinkAutomation implementation with failover. Please contact our Professional Services team for more information.