Standby Orchestrator setup

Help Orchestrator
,
1

Team,

We have a single node deployment in Production. We need to mirror the deployment to another server which can be named as Disaster recovery and booted in case Primary prod goes down.

Please note that we do not want to have High availability Add on configured.

Is it possible? If yes, can you please provide the steps.

Regards,
Ankit

2

Hi @Ankit_Aggarwal

The following doc puts light on the active/passive disaster recovery setup, it might help you with your requirement:

In addition, you can simulate the backup server architecture something like this:

  1. Install UiPath on the disaster recovery server, using the same version and configuration settings.
  2. Both the primary production server and the disaster recovery server must have access to the same resources (file shares, database servers, SMTP servers, etc).
  3. Configure the necessary settings on the disaster recovery server, including the licensing, certificates, and machine keys.
  4. Create a backup of primary production server’s UiPath installation and database, and transfer it to the disaster recovery server.
  5. Restore the backup on the disaster recovery server, overwriting any existing databases.
  6. Test the disaster recovery server to ensure that it is functioning properly and can handle the load of your primary production server in the event of a disaster.
  7. Set up a mechanism to keep the disaster recovery server in sync with the primary production server, such as using a scheduled backup and restore process or a replication tool.

Hope this helps,
Best Regards.

3

Hi @arjunshenoy,

Thanks for the detailed info. I have gone thru the document and that mentions about having a HA Add-on cluster but I want to set up a DR with single server deployment across primary and DR, so that means that I dont need HA cluster right and steps you explain would suffice.

Also, please help me if we can also setup a DR for the BOT machines as well:

Scenario is that there are few BOT machines and we want to setup a scenario wherein if the heartbeat for the BOT machines from orchestrator, it fail-over to the corresponding DR machine. Is it possible? If yes, can you please help in giving some details around that ?

Regards,
Ankit

4

Hi @Ankit_Aggarwal

Since your priority is to monitor the heartbeat, and fail-over to the DR without HA add on with few BOT machines being available, you can design the architecture something like this:

→ You can basically set up a heartbeat monitoring mechanism, such as a scheduled task in Windows Task Scheduler or a custom script, that periodically checks the status of the BOT machines in the primary environment. (Let’s say a PowerShell script can use the Orchestrator API to query the status of the BOT machines in the primary environment and check if they are responding.)

→ If the heartbeat monitoring mechanism detects that a BOT machine in the primary environment is no longer responding, trigger a script or a workflow that will stop the job running on the failed BOT machine and start the corresponding job in the DR environment. Note that this failover scenario will require some manual intervention to initiate the failover process, unlike the HA scenario where failover happens automatically.

Hope this helps,
Best Regards.

5

Hi @arjunshenoy,

can we have a quick call to discuss this? It should not be for more than 10-15 mins.

Regards,
Ankit