The term "disaster recovery" can initially make this entire class of essential IT backup and restoration services seem like something that is only necessary in the most exceptional circumstances. After all, what qualifies as a "disaster" and how much action is needed to respond to mistakes – such as having a secondary server go down – that might not seem that dramatic when they happen?
The outline of current disaster recovery attitudes and practices
As a result of this type of mindset, too many organizations are complacent or at the very least naive about their real disaster recovery needs. Data is routinely lost to technical glitches, not just to natural disasters that strike data center facilities or denial-of-service attacks that take down IT infrastructures for hours at a time. According to a recent TechTarget survey:
- Only 24 percent of companies have "high" confidence in their DR solutions. Nearly half (48 percent) had "moderate" confidence, while 28 percent had either "low" or "no" confidence in these tools.
- Steep technical challenges may be responsible for this lack of faith. More than half of DR customers need to recover applications from multiple physical locations and data centers. Most of them also have to duplicate processes that run on up to 99 physical and/or virtual servers.
- Such weak confidence may also stem from the fact that most DR systems are rarely if ever tested. More than 40 percent of them are examined only once per year; 22 percent are not put through the paces at all, meaning that it is literally impossible to know how they would behave in a real crisis.
- The vast majority of DR platforms do not cover all applications across the organization. Only 26 percent of organizations reported having comprehensive DR in place, while the rest had DR for mission-critical apps only or a group of applications that included some such essential programs.
- When evaluating DR services, prospective buyers give the greatest consideration to compatibility with specific virtual servers, backups and storage infrastructures, as well as ease of use and overall capacity.
To be fair, DR is a task with a scope that is often not well understood by IT departments until they actually experience a major incident. A 2014 Dimensional Research study, surveying 453 IT professionals in the U.S. and Canada, found that although 79 percent of respondents had suffered a significant outage within the previous 24 months, less than 10 percent reported that they would be able to effectively respond to such an event by recovering everything they needed within just 2 hours.
What are the most common obstacles to effective DR?
We have seen that setting up DR is not always easy, due to a combination of unfamiliarity/inexperience with the stakes of an outage and difficulty in ensuring that the proper range of applications is covered. Let's dive into a few of these challenges in more depth and what can be done to address each of them. Effective DR is often impeded by:
A DIY mindset
Organizations that manage DR on their own must shoulder the high burden of maintaining the relevant infrastructures. This responsibility entails significant capital expenditures toward equipment that is unlikely to be fully utilized, meaning that a sizable chunk of the investment is always wasted. Moreover, technicians must ensure that both primary and secondary DR sites are tightly synchronized, with the same configurations, drivers, etc. so that there is no significant drift between them that would jeopardize an actual DR operation.
An alternative to this approach is to work with a managed IT services provider and/or pursue Disaster-Recovery-as-a-Service. These options provide superior service-level agreements, reduced recovery time objectives and less overhead when compared to do-it-yourself DR.
"Some approaches to DR are penny wise but pound foolish."
Some approaches to DR are penny wise but pound foolish. For example, a DR solution may be predicated on cheap storage-area networks along with similarly inexpensive cables, disk drives and memory sticks. An organization that relies on such a set of infrastructures may save money in the short run, but lose more of it in the long run due to ineffective response when a disaster occurs. It is unwise to rely on a legacy DR system that seems to be fine but has never really been tested, either as part of a dry run or during an actual event.
Accordingly, it is a good idea to invest in DR solutions that are supported by Tier 3+ data centers and secured with cutting-edge technologies. Instead of sinking money into hardware and software that might not even work in a pinch, it literally pays to ensure that you have rock-solid infrastructure that gives you instant failover and peace of mind.
Improper scope for a business
There is no one-size-fits-all plan for DR. A small business is not going to have the same requirements – or the same financial and technical resources to draw upon – as a large enterprise. So it would not make sense for an SMB to saddle itself with a vast set of expensive IT infrastructures and lengthy pieces of documentation just to guarantee proper backup and recovery of its relatively limited assets. By the same token, a bigger organization cannot afford to skimp on DR, especially when it has multiple geographically dispersed sites that require attention.
What is really needed is a custom approach, under which the DR provider fully assesses each customer's requirements and devises a matching solution. For instance, whether to use virtual standby DR (which is highly economical) or replicated DR (which is better suited for mission-critical applications) will depend on a given company's structure and requirements.
UbiStor offers a diverse portfolio of DR solutions designed for a wide range of objectives. We work closely with each customer to make sure that the DR platform in question is properly scalable, cost-effective and dependable. Learn more on our virtual disaster recovery page to get started on the path to superior DR.