Converging Continuous Availability and Disaster Recovery with Just Two Data Centers – You Really Can Have It All

Converging Continuous Availability and Disaster Recovery with Just Two Data Centers – You Really Can Have It All

The gold standard for mission-critical business continuity is continuous availability and rapid disaster recovery, but many organizations struggle to achieve both, especially when application performance, tight budgets, limited staffing, and 

other challenges are added to the risk management equation. The traditional narrative has been that continuous availability and rapid recovery requires a three-data-center configuration or at least three database installations, but Axxana’s Phoenix overturns that story in an elegant, cost-effective way. Using Axxana’s Phoenix to converge continuous availability and disaster recovery solutions into one, organizations can significantly reduce the cost and complexity of business continuity by eliminating the need for one complete standby installation, including servers, licenses, communication ports, cabling, power, cooling, and so on. That is, instead of two complete standby installations, only one is needed.

The Challenge – Continuous Availability vs. Disaster RecoveryWorks.

Continuous availability and disaster recovery are two vital components of business continuity; however, each has different and contradictory requirements. To meet these requirements and provide a complete business continuity solution—that is, one that includes both continuous availability and disaster recovery capabilities, at least three database servers are required: a production server, a local standby server for continuous availability, and a remote standby server for disaster recovery. To understand why, let’s review some of the key aspects of continuous availability and disaster recovery solutions.

Continuous availability solutions guarantee zero downtime in a single-point-of-failure scenario.

They typically accomplish this by using synchronous replication to fail over (transparently to the applications) to a standby database server in the event of a failure at the primary database server. To ensure acceptable application performance, the primary server must be close to the standby server, as replication latency in a synchronous replication topology has a direct bearing on the latency of each transaction. The problem with this topology is that it is vulnerable to regional or rolling disasters. In addition, synchronous replication is extremely bandwidth-intensive, which creates high communication costs.

Disaster recovery solutions enable smooth failover after a major data center failure.

However, to ensure immunity from regional or rolling disasters that impact the primary data center, the standby server must be at a significant distance from the primary data center. The large distance between the two sites makes synchronous replication prohibitive (due to communications costs and performance degradation). Without synchronous replication, transactions will be lost in a failover. In addition, before failing over, organizations may have to assess whether indeed the data center can be recovered, and if so, how long it will take to recover the data from the primary site. In many cases, organizations decide to wait until the primary site recovers, as the risk of failing over and losing transactions is too high. Depending on the disaster, downtime may extend for hours or even days.

Many organizations compromise business continuity by choosing one or the other solution because they mistakenly believe that they cannot afford the cost and complexity of both. With Axxana’s Phoenix, such a flawed strategy is no longer necessary.

Axxana’s Phoenix: The Perfect Converged Solution

Axxana’s Phoenix, used with Oracle Active Data Guard, provides a perfect converged continuous availability and disaster recovery solution using only two data centers.

it significantly reduces downtime in case of a more severe failure incident. That’s because Axxana’s Phoenix ensures zero data loss. Failing over to the remote site—and in many cases, failback—is much less risky and costly; and therefore the decision to fail over is easier and quicker. In addition, the data reconciliation process is totally eliminated, saving both time and money.  

Given these advantages, there is no longer a reason to compromise. You really can have it all.

For technical details on how Axxana’s Phoenix enables both continuous availability and disaster recovery within a two-data-center solution, please see our white paper.