Disaster recovery (DR) is more than just the restoration and validation of data. It often involves the reinstallation applications and painstaking reconfiguration to reflect changes that have been made over time. In today’s highly complex and ever-changing IT environments, it’s difficult to keep track of those changes and incorporate them into DR plans. However, DevOps processes and automation can also be applied to DR, dramatically reducing the time it takes to restore applications and services.
Soaring downtime costs and more stringent service-level requirements have caused recovery time objectives (RTOs) to shrink. Effective DR planning is increasingly critical to the business, but is extremely difficult to get right. The DR plan for a given application is typically created when the app is put into production — after which, it is often forgotten. It isn’t kept up-to-date as patches and updates are applied to the application, which can lead to version conflicts and other issues during the restoration process.
Ongoing testing of the DR plan is another major challenge. Testing requires IT to run through the DR plan and ensure that a successful recovery is possible. It is a time-consuming, expensive and complex process that can disrupt the production environment. Because DR testing has the potential to impact customers and revenue, many organizations don’t test at all, while some test offsite using offsite backups. However, failing to properly test the DR plan can be a recipe for true disaster and extended downtime.
How can DevOps improve DR planning? DevOps enables continuous integration (CI) and continuous delivery (CD) of software changes so that applications can be managed more effectively and modified as needed. It also uses automated tools for application deployment and incorporates a process for validating applications between development and testing environments and production. This makes it possible to accelerate application rollouts while minimizing human error.
In other words, IT teams that use a DevOps approach regularly go through procedures for testing changes to applications. By incorporating DR planning into the DevOps pipeline, IT can ensure that the DR plan is managed along with the application.
The integration of DR into DevOps also acknowledges that recovery is essentially the same as an application deployment. If an application is implemented only once, the knowledge acquired during implementation will likely be lost by the time recovery is required. A DevOps approach enables IT to apply the ongoing deployment experience gained through CI/CD to disaster recovery processes. In addition, the automated tools that are used to move applications from dev/test into production and back again can be used to failover and recovery. DR environments, which replicate production, can be utilized for DevOps workspaces to make the most of capacity that traditionally sits idle.
DR planning has always been hard, and it has become more and more difficult due to growing numbers of applications and services and the accelerating pace of change. DevOps processes are designed to help IT better manage change, and can improve DR planning through ongoing updates, regular validation and automated tools.
January 8, 2019