Disaster recovery data center – basic technological solutions

27.03.2018
Outsourcing of IT infrastructure, Security
Author: Atman

Depending on the purpose to be met by the disaster recovery data center, there are several options to choose from. Between the colocation and the public cloud – the extreme solutions – there is an entire spectrum of intermediate solutions, including private cloud or dedicated servers.

The first option is the basic colocation service. It is already a well-known product on the market, but in the case of a disaster recovery data center, it is often enriched with dedicated links between the core and backup centers. In this configuration, the distance between the two data centers is really important. If the solution is to provide failover in the active-active mode, then the most common optical distance cannot exceed 30 km. Colocation is a solution chosen by clients who want to rely on dedicated equipment and – at least for now – exclude the option of using disaster recovery center solutions in the public cloud computing. Sectors such as banking and finance, insurance or state administration are particularly vulnerable to these issues.

The second option is to build a disaster recovery data center using a public or private cloud. This solution is intended for companies looking for redundancy with the least operational involvement. There is no need to communicate data processing centers using dedicated telecommunications links or employ specialists maintaining IT infrastructure. The solution has a low entry barrier and the ability to run virtually overnight. Its disadvantage is that not all redundancy scenarios are possible to be implemented, while with more complex systems (especially legacy) architecture is to be reengineered. It may also turn out that in this way we will be able to secure only part of the data processing processes. In most cases, the public cloud is a good choice, and for backup of important data, probably there is no better solution, if we take into account the flexibility and financial expenditure that must be incurred in order to use it.

Those clients who expect a more individual approach are more likely to choose a private cloud, which is a tailor-made solution, prepared to meet very specific requirements. For such clients, the lease of dedicated servers is also a good solution.

The third option is a hybrid solution, which also assumes the support of the organization at the operational level. Regardless of the selected disaster recovery data center architecture (colocation, public cloud, dedicated hosting, private cloud or mixed model), the client also gets access to backup offices. This is an extremely important service, taking into account the need to maintain business continuity. In practice, even the best technical solutions are unhelpful if a person is excluded from activities. The need for access to such back-up offices results not only from the need of protection in the event of a natural disaster (fire, flood) or a terrorist threat. The inability to carry out tasks often results from more prosaic reasons, such as heating failures in the winter season or interruptions in access to electricity. This is well illustrated by the research we carried out two years ago. It turned out that more than half of the companies surveyed by us experienced problems with access to the Internet or electricity. As a result, 41% of them had to put on hold or seriously reduce their activity. Interestingly, most often they were not short-term incidents. Only in 6% of cases the failure lasted less than an hour. In 30% of cases, it lasted from 1 to 2 days, and in 10% of cases even longer. It is not difficult to count losses in such a situation.

The synergy resulting from having both a disaster recovery data center and backup offices at one provider means similar procedures. It is also often one logistic place, easily accessible, where we can recover our key business processes.