Architecting The Perfect Continuous Availability And Disaster Recovery Environment

That is to say, what is your stated service level agreement (SLA) for application and data recovery time and recovery point objectives (RTO and RPO)?

RTO deals with the how fast an application has to be back up and running after a failure, while RPO deals with the volume of data that can be lost. RTO and RPO can also be defined in terms of capital. That is, how long can you afford to be down and the amount of data can you afford to waste. When defined in monetary terms, it becomes rather clear as to where the focus needs to be.

There are numerous solutions that provide data RPO in the seconds range. However, by comparison, there are surprisingly few solutions that can also deliver the same, for application RTO. Neverfail is one of the select few, to have developed the required technologies required to deliver a solution that provides application RTO and RPO in the near zero range, quickly getting rid of application downtime.

So, how do you know which applications are the most important?
Which require RTO/RPO in the near zero timeframe?

The solutions to such questions are in fact pretty simple. Which applications bring about the most help desk calls when they are unavailable? Which applications, when unavailable, cause the organization to lose money, or productivity, or customer confidence?

Making these decisions and being alert to the affect of application downtime will help you to decide which applications require the most attention. Establishing which applications require the most precise RTO and RPO is the first step in deciding what type of solution and architecture will work best for your business.

Over the years, a vast array of point products has been developed to solve the specified availability problems of the time. Whilst there has always been the need for a solution to bring applications back online quickly after a failure, with as little loss as possible, discovering a complete solution to the problem has remained elusive. However, more recent technology advances (in replication, clustering, e duplication etc.) mean that is now possible to move away from the well known Recovery Centric approach to clearing up the situation, towards a more Availability Centric approach.

Even though choices have been made with regard to the criticality of applications in an environment, there is still an overriding fact that cannot be excluded, application downtime is intolerable for most application owners, users, customers and management. Given this, it seems like ensuring the application is adequately protected and monitored, so that a pre emptive action can be taken to solve a issue before it becomes an concern, is key.