The importance of disaster recovery (DR) is well established,especially with enterprises that use critical applications. After doing DR the traditional way, enterprises are now leveraging the concept known as disaster recovery as a service.
The traditional way of DR implementation includes setting up one's own site with the required IT infrastructure, and then managing it as effectively as the primary site. However, there is growing concern about the direct and indirect costs associated with this option.Further, the resources at the DR site remain idle most of the time.While it's certainly good to have a DR site to cover any disaster thatmay strike a company, one question which arises is whether there's a more cost-effective alternative.Disaster recovery as a service attempts to answer this question, giving organizations an opportunity to optimize costs on many accounts.
Another major advantage of using disaster recovery as a service is the availability of an alternate human resource pool (say, system and database administrators) from the service provider (SP). This factor is important should the primary team get into a state of trauma as a result of the incident that led to the disaster. SPs can also use their professional experience to add value during the drills to enhance the DR preparedness of client organizations.
While evaluating disaster recovery as a service, there are some aspects which a company should check internally, as well as other factors which it should check with the SP. Internally, as the first checkpoint, a company should find out whether its own data security policies comply with regulatory requirements. If this isn't an issue, it should then assess the TCO of maintaining dedicated DR infrastructure for itself, so that a comparative study can be made.
With regard to the SP providing disaster recovery as a service, the company should check the portfolio of services being offered by the vendor. It is advisable to check the competency of the SP to bring the company's systems to at least a warm state of operations, should the DR be invoked. Both the company and SP should work together during periodic drills to build such competencies.
After selecting a vendor to provide disaster recovery as a service, the next step is to decide the service level agreements (SLAs). Some of the key specifics that SLAs should include are:
- Lead time to allocate the minimum required resources, should DR be invoked.
- Lead time to scale up resources to the defined (or full) level.
- Duration for which such resources will be retained on a dedicated basis for the company.
- Additional fees for occupancy beyond the pre-defined period.
- Additional facilities such as conference rooms and video conferencing.
- Capability to provide additional hardware as and when needed.
- Parameters related to work area recovery can also be included if such services are used.
Now, what are the issues associated with using disaster recovery as a service? One of the common concerns with disaster recovery as a service is the multiple simultaneous invocations which lead to resource contention. Based on the SP's size and capacity, this issue can be resolved if some base rules are set. In a simple scenario, problems could be solved on a first-come-first-served basis.
About the author: C Kajwadkar is currently the chief architect and vice president of Netmagic Solutions.
(As told to Anuradha Ramamirtham)