Business Management

Rob Ciampa (UK) - Business Continuity Means Not Accepting Failure

Mathematician and author, John Allen Paulos, is noted for advancing numeracy through his books, lectures, appearances on popular American television programs and the "Who's Counting" column on ABCNews.com. He is also the source an oft-used quote: "Uncertainty is the only certainty there is, and knowing how to live with insecurity is the only security." I understand the risks of contradicting a scholar with impressive credentials, but Dr. Paulos is only half right; at least when his somewhat fatalistic perspective on living with insecurity is applied to business continuity. Let me explain.

Wikipedia defines business continuity as "the activity performed by an organization to ensure that critical business functions will be available to customers, suppliers, regulators, and other entities that must have access to those functions." I'd expand on this definition by adding that it is incumbent upon those responsible for business continuity to guarantee that key operations of a firm continue without stoppage, even in the wake of the adverse circumstances or events.

I'll concede an IT bias given my chosen profession, but I'm reasonably confident that many IDG Connect blog readers will agree that protecting an organization's information technology systems is an immensely critical component of maintaining business continuity; perhaps the most important. The fact of the matter is that when IT stops, business stops. Unplanned application downtime can harm an organization's brand, productivity, and relationships. These are bad outcomes that result in lost revenue and business inefficiencies. IT needs to deliver constant application uptime; it must prevent, not just react to, downtime.

Application availability is the discipline of employing best practices to enable continuous, uninterrupted access to the systems and software. The time has come to view application availability as a strategic requirement. I'm encouraged by recent conversations I've had with Marathon Technologies' partners and prospects across the globe. Executives, line-of-business managers, business process owners and, yes, IT professionals from DBAs to CIOs have embraced the reality that application availability is a business imperative, regardless of whether their business is small, mid-sized or a multinational conglomerate.

This anecdotal evidence is unequivocally supported by industry analyst research. For example, this past September Forrester Research published a report titled "Business Continuity And Disaster Recovery Are Top IT Priorities For 2010 And 2011," (Stephanie Balaouras, et. al., September 2, 2010). Forrester surveyed more than 2,800 IT decision-makers to determine top IT budget priorities. The study found that business continuity topped the list of priorities for SMBs and was second only to consolidating infrastructure among respondents with large enterprises.

If you have a role in maintaining business continuity for your organization, living with insecurity should keep you awake at night. And it is certainly anathema to business continuity planning. My suggestion is to prevent failure by planning to avoid it from the start. This advice may sound simplistic, but when you consider that far too many organizations have inadequate measures in place to ensure availability of databases, servers and the business applications they run - and often none at all - the point is worth consideration.

Traditional approaches to application availability do not meet the requirements of the new always-on world because they are reactionary methods that are satisfied with speeding up recovery times after a failure rather than preventing downtime in the first place. Two of the more common of these reactive approaches have been replication with manual failover and clustering. Both are considered high availability (HA) solutions. And each has non-trivial shortcomings. For starters, both are "fail and recover" techniques that concede downtime. Choosing clustering or replication means acquiescence to falling short of the business requirement for always-on application availability. If that's not enough reason to consider other availability options, factor in the cost and complexity of a shared storage network (SAN) required by a cluster and the need for manual intervention by IT staff. Both approaches are prone to disproportionate responses to minor issues, i.e., network interface or RAID controller failure, the entire systems fails over to the second server. Not only does it take time for the backup server to spin up, but until the issue with the first server is mitigated an organization is vulnerable because it has a single point of failure. If "usually" or "mostly" available applications won't cut muster within your organization, HA is not the right choice.

There is another option. Fault tolerance (FT) provides the highest level of application availability because it allows systems to compute through failures without interruption. Historically, this guaranteed availability has come at a price. FT hardware is expensive and requires considerable services to configure and support. This means that many proposed FT projects are killed by the cold calculus of budget reviews. Then there is the dreaded vendor lock-in. Organizations settled for an HA solution or overspent on hardware-based FT.

FT software is a more attractive option. Choices include purpose built solutions from companies like Marathon Technologies. Some may consider virtualization a path to availability, but it fundamentally lacks the ability to ensure no loss of compute cycles and critical data. Availability requirements need to be separated from consolidation requirements.

There are software-based options that provide flexibility and deliver FT performance at reasonable costs, especially when compared with hardware-based FT solutions. When considering FT software, consider products that address both symmetric multiprocessing (SMP) and multi-core servers and applications. Doing so will ensure application availability and compatibility.

Business continuity means never planning for failure. Start by identifying core applications and databases that absolutely, positively must always be accessible. These may vary based on a business or industry but probably include customer-facing applications (for example, CRM), transactional systems (such as ecommerce), order processing and supply chain management. Once this list has been developed, select the right technology for the job. Proven FT software isn't expensive. Business outages are.


Rob Ciampa is VP Marketing for Marathon Technologies, a business application availability software provider based in Massachusetts, USA and has offices in Bracknell, UK.





« Axel Pawlik (Middle East) - The Future of Internet Development in the Middle East


John Goldrick (Australia) - 2011 Predictions: Mobility and the Unwired Enterprise »


Do you think your smartphone is making you a workaholic?