Application Availability

Apart from protecting your data, another cornerstone of the disaster recovery plan is figuring out how to make sure your important applications can be kept running or recovered as quickly as possible in case of disaster.

Email servers, ERP systems, trading platforms, and similar critical applications often need to be recovered within minutes or at most a couple hours of a failure. What technology you choose to do this depends on a few factors.

Risk: Make sure you understand what kind of risk you are trying to protect yourself from. If it’s hardware failure, then a fault tolerant system or onsite redundant hardware may be sufficient.  If it is a site failure due to fire or other disaster, then having backup servers offsite is required.

Recovery Time Objective (RTO): The basic question you are asking yourself here is, “how quickly must I be able to get my application back online after a disaster happens?” The answer usually depends on the business impact the loss of that data has on your critical business functions.

Recovery Point Objective (RPO): The basic question you are asking yourself here is, “how recent does the data on my backup server have to be?” For some applications, like online banking, the answer may be milliseconds. For other applications, such as email, maybe a few minutes old is ok.

Budget: If money is no object, then real-time mirroring of data using dedicated fibre optic connections between two data centers at least 10km apart is the gold standard. This is what some banks do. If your budget is less expansive, then asynchronous replication to an offsite location can provide near-instantaneous failover.

For more information on application availability, please read about the solutions we recommend, or contact one of our consultants.