Managing Load Balancers During a Recovery using OCI Full Stack DR

Full Stack DR is an (OCI) disaster recovery orchestration and management service that provides comprehensive disaster recovery capabilities for all layers of an application stack, including infrastructure, middleware, database, and application.

FSDR Service currently supports disaster recovery for the following OCI resource types:

• Compute Instances.
• Boot and Block Volumes (Volume Groups).
• Oracle Exadata Database Service.
• Oracle Base Database Service.
• Oracle Autonomous Database Serverless.
• File Systems.
• Load Balancers.
• Network Load Balancers.

The load balancer feature was recently added to the support list above, and in this blog, I will explain how to manage a load balancer during the FSDR process.

Prerequisite

In this blog series, I won't cover FSDR creation and will focus on managing the Load Balancer during the disaster in part 1. I will add the custom hostname to the LB and create the traffic steering policy to manage the load balancer DNS during the DR in part 2; however, you need Knowledge of OCI networking, Load Balancer, DNS solution, and Full Stack DR solution.

You also need to have the created resources listed below.

• FSDR solutions with Disaster Recovery Protection Groups (DRPG) in both primary and secondary regions.
• Virtual Cloud Network (VCN) with subnets, proper security list, and Internet Gateway in both regions.
• Load balancer in both regions.
• Compute instance.
• Volume Group for compute instance Boot volume.

For more information about how to create FSDR, please check this link out. 

Agenda

• Add moving instance, volume group, and load balancer resource members to DRPG primary region San Jose. (Part-1).
• Add load balancer resource member to DRPG secondary region Phoenix.
• Create a Switchover plan from the secondary region Phoenix
• Execute DR plan in Phoenix.
• Validate the new members in the new primary region, Phoenix.
• Create a traffic management steering policy with a custom load balancer hostname (Part-2).
• Execute the DR plan to validate the load balancer failover.

Add moving instance, volume group, and load balancer resource members to DRPG primary region San Jose

Log into the OCI dashboard, and from the hamburger menu, click Migration & Disaster Recovery, and then navigate to DR Protection Groups.

Under resources, click on members and add members.

Choose Compute from the available resource type, available compute instance in a San Jose region, and click on Moving instance type. 

Now click on Add VNIC mapping. 

Note: Moving instance will move from the primary DR protection group to the standby DRPG during DR operation.

In the Add VNIC mapping page, VNIC is the compute instance(FSDR-SanJose-VM) available in a San Jose region; Destination VCN(Demo-VCN-PHX) and subnet( avaialble subnet in Demo-VCN-PHX VCN) exist in a secondary region.

Now, add the next member, Volume group.

Volume group simplifies creating time-consistent backups of running applications. The existing compute instance (FSDR-SanJose-VM) in a primary region, San Jose, has a Boot volume already added to the Volume group (Demo-VG); otherwise, you will get a failed state during the member creation. 

Check the screenshot below.

For the last member in the DRPG primary region, I added the OCI load balancer, which the compute instance added earlier is its backend server.

Click on Add member, choose the resource type load balancer, and click on available LB in your region.  

Next, you will add the destination load balancer name and backend set.

Note: The source backend set is San Jose load balancer BS with FSDR-SanJose-VM (compute instance added earlier) as a server. 

Now check the DRPG details page below.

Check the Disaster Recovery Overview page and confirm the region's status

Add load balancer resource member to DRPG secondary region Phoenix.

Change the region from the right side of the OCI portal page to the Phoenix region and navigate to the Disaster Recovery DRPG detail page.

Due to the creation of moving instances in the primary region, you don't need to add compute and volume group members here.

However, adding the existing load balancer in the Phoenix region is mandatory; otherwise, you will get a failed state.

Add load balancer as the only member in the Phoenix region. Check the screenshot below. 

Check the details page below.

Execute DR plan in Phoenix.

Primary and secondary DRPG resources are ready to test the DR plan now.

From the DRPG details page, click on Plans to create a Switchover plan.

Check the screenshot below.

After creating the plan, click on the Execute DR plan. Check the screenshots below.

Click on Execute DR plan and navigate to the plan execution groups detail page.

This process is one of the critical steps with the FSDR solution, which lists all the step-by-step processes, from validating the resources to moving them to the DR region.

You need to wait until all the processes state change to succeed.

Note: Terminate Compute Instance and Volume Group are disabled by default, but you can enable them based on your requirements.

After the execution plan is finished successfully, you will see San Jose moving resources added as the new members to DRPG for the Phoenix region.

Check the Phoenix LB with the new backend server below.

Below are the details of the instance FSDR-SanJose-VM initially created in San Jose with the new IP address assigned from the existing VCN in Phoenix.

To continue,  check the second blog.