Real-time RPO replication

Details about real-time data replication in Arpio

When configuring Arpio to create recovery points for your application, you have the option of enabling  a recovery point objective (RPO) of "Real-time replication".

When real-time replication is enabled for an application, all EC2 instances in that application will have a near real-time recovery point objective of seconds. Arpio achieves this by leveraging an integration with AWS's Elastic Disaster Recovery Service (DRS). 

This article will provide an overview of real-time replication in Arpio, as well as address some frequently asked questions. For documentation on how to enable real-time replication in Arpio, click here.

Understanding your real-time recovery-point objective (RPO)

Your recovery point objective is defined by the maximum gap of time your business application can afford to be down before service is restored. When real-time replication is selected for a resource, your RPO will be reduced to seconds.

See Arpio’s documentation for more details about your application’s RPO in Arpio. 

Replication lag

While unlikely, real-time replication lag can occur in specific scenarios, and your RPO can increase beyond seconds. These environmental factors are out of Arpio's control and will likely correct themselves. 

Snapshot policy for resources with real-time RPOs

AWS DRS maintains an aggressive point in time (PIT) policy for snapshots, which is automatically set by AWS DRS. Those policies' defaults are:

  • Every 10 minutes for the last hour
  • Once an hour for the last 24 hours
  • Once a day for the last 7 days*

*Arpio will calculate this retention policy based on your Arpio application Recovery Point Retention and there is no further action needed. 

Real-time resources in your recovery environment with non-real-time resources

Today, Arpio supports real-time replication for a EC2 resource types only. When you have real-time replication enabled for an application that includes other resources that are not real-time replicated, Arpio will apply the lowest RPO it can to them. Usually around 15 minutes.

Replication vs. Snapshots

There will likely be servers in your source environment that are not critical to recovering instantly if a disaster strikes. You still need them in your recovery environment. Arpio recognizes this and will apply the lowest RPO available via low-cost EBS snapshots on these servers that do not have the AWS Replication agent installed. More details about how this set up works can be found in our documentation Enabling Real-time Replication.

Cloud costs associated when using Arpio with the DRS service

Arpio uses integration with AWS’s Elastic Disaster Recovery service to achieve real-time replication. This will introduce additional costs to your AWS bill that include the agent, replication servers, and snapshots. Please review Amazon pricing for DRS with examples

Note: Costs associated with cross-region replication are not detailed in the Amazon pricing for DRS. These cross-Region data transfer costs are outlined here. EC2 on-demand instance pricing can be found here.