The 2008 hurricane season displaced millions of Americans across the Gulf states and eastern seaboard and ground commerce to a halt in Houston, the country’s fourth largest city for more than a week. Hurricane Ike reportedly caused anywhere from $27 to $52 billion in damage in Texas alone, while serving to remind companies they are only as sound as their business continuity plans.1 It may not be possible to completely disaster-proof a company against a Category 4 hurricane or other major natural disaster, but shoring up your company’s physical infrastructure layer will provide a solid foundation for any business continuity plan.

 

Until recently, the importance of the physical infrastructure layer (i.e., the cabling and interconnections which are part of layer 1 of the Open Systems Interconnection [OSI] model) has been largely taken for granted. Natural disasters have served to increase awareness that realistic business continuity plans and testing methods are imperative. Still, hardware and cable failures remain the leading cause of downtime. After you’ve secured appropriate generators, stashed reserves of gasoline and stocked up on food, water and other emergency supplies, it is imperative to focus on the undergirding structure of your company’s IT backbone. Without a solid, self-healing physical infrastructure layer, even the best-laid business continuity plans are likely to falter in critical times.

 

Unfortunately, natural disasters are not the only challenge facing American enterprise. According to a recent survey by Envoy World Wide, companies named the following as the top threats to business continuity: data security failure, including virus, denial of service or unauthorized access (62 percent), and data center hardware/software failure (58 percent).2

 

It is a common misconception that the physical infrastructure layer is relatively disaster- and hacker-proof. In reality, human error can unwittingly produce a security breach. A single misplugged cable, for example, has the potential to transfer highly confidential data outside the data center’s firewall. CSI statistics show that up to 80 percent of data theft and security breach incidents occur at the physical layer.3 In addition to physically securing network connections, zoning and logical partitioning of the network can also be implemented in a more robust fashion at the physical infrastructure layer.

 

The physical infrastructure layer is also stressed and vulnerable every time new equipment is introduced, old equipment is removed or existing equipment and cables are moved at a data center. Tangled and fragile cables, plugs and wiring can physically limit the ability to scale and consolidate - and most of these physical interconnects are not monitored for health and usage. As much as 70 percent of network downtime can be blamed on cabling problems.4

 

To facilitate both business continuance and disaster recovery testing and take the data center to of next level the automation, the physical connectivity infrastructure layer must be addressed. The best systems can provide connectivity when and where it is needed - with provisioning, monitoring and protection at the click of a mouse. Regardless of the mix of IT equipment and network protocols and interfaces, data center networks require a flexible and adaptable physical infrastructure layer to physically interconnect devices. A dependable and malleable physical infrastructure layer allows for automated and secure equipment moves, adds or changes, thus enabling seamless reconfiguration and testing conducted in accord with the dynamic needs of a virtualized data center. Without such testing, there is no efficient way to judge the effectiveness of your company’s business continuity plan until it is too late. An automated physical infrastructure layer provides a more secure environment by enabling a hands-off approach to managing the network. This remote access is particularly valuable in the event that IT personnel must be evacuated from a data center, or if the building becomes unreachable due to flooding or other impediments. Finally, an automated physical infrastructure layer enables improved network resiliency that could help to bring operations back online faster in the aftermath of a disaster.

 

When natural disasters are not a concern, this robust physical infrastructure layer offers improved monitoring capabilities to quickly identify data security breaches and other network issues. Typically, standalone diagnostic tools are connected in line with problematic paths in the network, but this approach makes for an inflexible model that is difficult to expand when your business grows. Using an intelligent infrastructure switch provides a much more integrated and scalable solution to monitoring and diagnostics. For example, administrators have the ability to route any path(s) within the infrastructure passively into a desired test device without ever moving a physical cable or changing its end-to-end circuit characteristics. Further, intelligent infrastructure switches provide increasing levels of diagnostic and tracing facilities that are integrated into the product and act quickly to expose and enable correction of network problems.

 

Whether your company faces a natural disaster, the trials of the test lab or the daily grind of e-commerce, enhanced diagnostics will help to scan data paths and capture protocol traces automatically on error conditions. An intelligent infrastructure switch has the ability to test network reaction to loss of signal by issuing a test command. Additionally, an intelligent infrastructure switch merges network performance monitoring and diagnostics with automated connectivity functionality. Management interface tools available in these systems can be better used to provide visibility into the network infrastructure, allowing for more accurate event correlation as well as providing alarms, alerts and diagnostic information in multiple formats, including SNMP and email home capabilities.

 

No matter your operation or its exposure to Mother Nature’s wrath, today’s enterprise-class IT equipment guarantees a fault tolerant architecture providing at least “five nine’s” (99.999 percent uptime) availability. This requires that each component in the network have no single point of failure (redundant AC power, DC Power, functional failover/failthrough, etc.) - and it all begins at the physical layer. Data center designers often forget that physical layer infrastructure failures are among the most common problem areas. Components including cables, patch panels and optical interfaces are among the most frequent failures in the network, resulting in points of vulnerability that expose the weak link in the network chain.

 

Therefore, it is important that resiliency is built into the connectivity infrastructure layer to help ensure maximum availability. To be robust, the physical layer of the network infrastructure must be designed to include features such as N+1 power and 2x capacity for redundancy and nonblocking. Administrators must also be able to configure failover paths for critical data connections. If an infrastructure component, such as a cable, fails in this architecture, operations will automatically be routed alternately and continue to function seamlessly.

 

In a fiber-based network environment, signal strength and signal degradation are a common cause of intermittent problems and can be difficult to isolate - and a real headache when your IT team is struggling to plug 10 holes at once. There will always be signal attenuation with each patch connection and bend in an optical cable, but an inherent benefit of switching at the infrastructure is that signals are monitored and redriven at each port. This enables a more resilient physical layer in the data center. An intelligent infrastructure switch can also be leveraged to extend distance via long-haul small form pluggable (SFP) optics.

 

Finally, a dependable physical layer facilitates the business continuity and disaster recovery testing that many businesses are required to conduct on a quarterly or annual basis. Even if your particular sector does not require such testing, isn’t it better to be safe than sorry? While some firms are loathe to schedule complete analyses (due to the volume of equipment and time required to test through traditional means), a properly wired infrastructure layer is able to run these critical tests at the click of a mouse. This method trims testing time, reduces system wear, saves man hours and eliminates the need to unplug/replug cables.

 

As evidenced by the recent demands of the 2008 hurricane season, today’s enterprise data center customers demand highly available, fault tolerant, scalable connectivity infrastructures to support their mission-critical LAN, MAN, SAN and WAN network environments. One of the best ways to fortify this complex system and ensure its continued secure operation and reliable uptime is to build up from an automated physical layer foundation known as a virtual infrastructure layer. Quantifiable in many ways, the net benefits include significantly improved network security, reduced time to implement disaster recovery tests, faster troubleshooting, and enhanced resource utilization and visibility.

 

Few executives may be willing to stare down a Category 5 storm when next year’s hurricane season rolls in, but life without a functioning network or data center invokes sheer terror for even the most daring members of the C-suite. Taking steps now to bolster your physical infrastructure layer is the surest way to prevent such a scenario.

 

References:

 

  1. Stewart Powell and Bennett Roth. “Texans Seek Aid in Congress as Ike Cost Estimates SoarHouston Chronicle, September 23, 2008.
  2. Envoy World Wide.“Trends in Business Continuity and Risk Management White PaperIDS-Emergency Management, October 29, 2008.
  3. Computer Science Institute. “The 12th Annual Computer Crime and Security Survey.” CSI.com, 2008.
  4. Ely Cohen. “It All Begins with the Physical Layer.” BICS News, May/June 2005.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access