With cable forming such a critical part of the network infrastructure, it's important to understand what types of cables are used for connecting Fibre Channel devices to a SAN. In the second part of a two-article series, Mike Harwood takes an in-depth look at fiber optics and discusses why fiber cabling has become the media of choice for SANs.
Copper-based cable has long formed the backbone of networks and over time has proven itself as both a reliable and easily managed media. Today's modern network environments, however, with their intensive high-speed data needs, are finding the limitations of copper-based cabling far too restrictive. Fiber-optic solutions are replacing copper infrastructures in many network applications, and as far as Fibre Channel-based networks are concerned, fiber cabling has emerged as the undisputed media of choice.
The advantages of implementing a fiber-optic cable solution are well documented. Simply put, fiber-based media does not suffer from the same restrictions and limitations of copper-based cable. Namely, it's immune to the effects of electromagnetic interference (EMI), radio frequency interference (RFI), and voltage surges. This is particularity important when it is necessary to run cables near electrical hardware, making fiber well suited for intra-floor conduits and for wiring duct spaces that often run close to florescent lights or other sources of RFI and EMI.
In addition, fiber optic cable is far less susceptible to signal loss (attenuation), enabling it to carry data signals significantly farther than its copper counterpart. While fiber cable does suffer from attenuation to a lesser degree, as with copper cable, a repeater can be used to boost signal integrity and increase the distance the cable can be run. In practical application, copper media may need repeating devices every 100 meters, whereas fiber cable may not require a signal boost until after a few kilometers. This means that along the same length of line, fiber-optic cable requires fewer boosting devices than copper cable.
While the advantages of implementing a fiber optic solution are clear, so too are the disadvantages. Unfortunately, a fiber-optic system is largely incompatible with the existing hardware infrastructure of many organizations. This deficit forces those organizations that wish to implement fiber optics to retrofit the current network infrastructure to accommodate a fiber-optic network. This is a costly endeavor. Furthermore, fiber cable is more difficult to physically install than copper media and often requires trained personnel and specialized tools.
Page 2: Getting Physical with Fiber Optics
Getting Physical with Fiber Optics
Despite the drawbacks of a fiber-optic solution, the performance gains achieved with fiber optics make it well worth the investment. To really understand why fiber offers such increased performance over copper-based media, it's necessary to take a look at the physical construction of fiber-optic cabling.
It is sometimes assumed that fiber-optic media, being glass, is quite fragile and must be handled delicately. While it is true that the fibers are made out of glass, optical fibers are a lot stronger than you might think and, once enclosed in a protective casing, can rival the strength of even copper-based media.
The strength of fiber cable is achieved using three separate layers. On the inside is the core, which is the central piece of glass tube used as a pathway for the light signals. Surrounding the core is the cladding, a type of sheath comprised of multiple layers of glass. This glass is used to reflect light signals back to the core, preventing light from leaking out. Both of these layers are encased in a buffer, a layer of hardened plastic that protects the core and the cladding. There are two types of buffers used in fiber optic cable, tight and loose buffers.
In a tight-buffered cable design, the buffering material is in direct contact with the fiber. Because the buffer layers are in direct contact with the fiber, any stress applied to the outside of the buffer is transferred to the fiber core. Tight-buffer configurations are generally used with indoor cables.
Loose-tube buffering systems separate the fiber from the buffer to minimize stress transfer. A gel layer between the fiber and buffer absorbs shock and high impact stresses. Cables incorporating loose-tube buffering system are much larger than those with tight-tube buffers and are often difficult to terminate. The construction of loose-fitting cables makes them well suited for outdoor application. Shown below is a comparison of tight and loose fitting fiber optic cables.
Tight-buffered Fiber Cable
Loose-buffered Fiber Cable
Page 3: Getting Physical with Fiber Optics
Inside the Core
The inside core of the fiber optic cable varies in size. The size of the core determines the type of optical cable — either single-mode or multimode cable. These modes describe the way in which light travels within the cable. Light signals can propagate through the core of the optical fiber on a single path (single-mode fiber) or on many paths (multimode fiber).
Of the two, multimode fiber optic cable has a larger core that allows for multiple streams of light signal to pass through simultaneously. These numerous light rays within the cable bounce around inside the core as they travel toward their destination. When light beams reflect off the sides of the core, they slow down and suffer from a reduction in strength.
There are two different types of multimode fiber cables, 50/125 micron and 62.5/125 micron. The 50 and 62.5-micron measurements identify the core's diameter, and the 125 micron size refers to the cladding diameter. The maximum transfer distance for 50/125 micron fiber is 500 meters, while the maximum transfer distance of 62.5/125 micron fiber is about 250 meters.
Both types of multimode fiber types offer data transfer rates of 133 Mbps. The smaller the diameter of the core, the farther a signal is propagated, as it uses a more focused path for the light rays to follow.
Single-mode fiber cables use a very small, focused core through which light can travel. This smaller core allows light to pass directly through the cable without bouncing off the walls. Single-mode cable has a center core of 9 microns and a cladding diameter of 125 microns. This smaller core provides for a maximum distance of 10 kilometers and a data transfer rate exceeding the Gigabit barrier.
Making the Connection
When it comes time to connect fiber-optic cable to the network, there are two types of fiber connectors used most often — 568SC connectors and optical GBICs. 568SC connectors are a color-coded duplex connector (beige for multimode and blue for single-mode). 568SC connectors are keyed to prevent incorrectly connecting the connectors. SC connectors use a push-pull design to mate and unmate a connection.
The other common connector type is the optical GBIC connector. Optical GBICs are used to convert optical signals to electrical signals and vise versa. There are two distinct types of optical GBICs available, shortwave and longwave. Shortwave optical GBICs are used with multimode fiber-optic cable, while longwave GBICs are used with single-mode fiber.
At the end of the day, there is little doubt that fiber optics has become the industry standard for terrestrial transmission of telecommunication information. The bandwidth needs of today's organizations simply require a medium that can deliver large amounts of information at fast speeds. While fiber solutions may be more costly and difficult to implement, it seems unlikely that copper cable will provide for future bandwidth needs.
See All Articles by Columnist Mike Harwood