by John Amato, Vice President, Data Center Systems
IT must support a growing number of applications and services running on a 24x7x365 basis and secure them against an array of constantly evolving cyber threats. Worse yet, IT organizations aren’t being resourced in proportion to their growing challenges, so it all must be done with unprecedented efficiency.
Just listen to three IT professionals talk about modern data center issues on a short video hosted on the interactive Fibre Channel Industry Associations’ (FCIA) Knowledge Vault.
IT leaders have embraced a variety of solutions to address these challenges, but they often overlook a critical aspect of their data center: the connectivity management systems that support mission-critical SAN, mainframe and network environments.
Data center connectivity is the equivalent of the human body’s circulatory system. It’s the lifeline between the “critical organs” of IT: the network switches, servers, director-class switches and storage devices. If this circulatory system doesn’t meet the demands of the vital organs, the health of data center IT as a whole is jeopardized.
Neglecting data center connectivity management leads to issues that can seriously undermine the health of the data center. These include:
• Inadequate performance. At higher speeds, connectivity infrastructure becomes far more sensitive to issues such as distance and imperfect connections leading to a loss of signal strength. So as tempting it might be to invest in to invest in 16Gbps Fibre Channel switch technology, it simply doesn’t make sense when the fiber plant itself can’t support 16Gbps.
• Cabling chaos. Fiber is becoming more prevalent and connections in the data center are often installed haphazardly over time. This makes changes and troubleshooting more time-consuming and susceptible to human error.
• Link-loss budget. High-throughput applications exhibit a lower tolerance for link loss than older IT services. IT has to optimize fiber connectivity attributes such as link distance, connection types and the total number of end-to-end connection and insertion points to ensure reliability.
• Termination and connection issues. Certain types of fiber connectors are especially susceptible to subtle reflection and misalignment impairments that can adversely affect high-performance applications and be difficult to diagnose.
Fiber Connectivity Management Best Practices
A set of proven solutions is available for transforming data center fiber infrastructure in ways that align with core IT objectives. These solutions include:
• Structured cabling. TIA-942 cable infrastructure design standards provide practical guidelines for key fiber connectivity parameters, such as cabling distances, pathways, site selection, space, layout and replication of all active ports on a CPL (Central Patching Location).
• Best-in-class hardware. Fiber hardware solutions can ensure the integrity and manageability of data center connectivity such as advanced “mimic” panels that prevent excessive handling of sensitive connectors, staggered cabling architectures that eliminate slack and factory-terminated connections that safeguard signal integrity.
By taking advantage of these solutions, IT organizations can upgrade to the high-performance, change-friendly fiber environment necessary to support intense data center workloads. They also can ensure that their data center fiber connectivity won’t cause performance bottlenecks.
The right fiber infrastructure to support evolving storage requirements protects the health of the data center connectivity infrastructure and empowers IT to:
• Better meet the escalating demands of the business
• Substantially reduce risk
• Improve allocation of limited OpEx budgets
Only fiber delivers the high throughput and low latency required by massive data applications. That’s why IT leaders should make a sober assessment of the current state of their data center’s circulatory system and take steps now to safeguard its health not just for the next weeks or months — but for the highly challenging years to come.