A large summary of fiber switch types
Entry level switch
The entry-level switch application is mainly focused on 8 to 16 ports of small workgroup, it is suitable for low prices, rarely need to expand and manage the occasion. They are often used instead of hubs, providing higher bandwidth than hubs and providing more reliable connectivity. People generally do not buy entry-level switches separately, but often buy with other levels of switches to form a complete storage solution. Entry-level switches provide a limited level of port cascading capability. If users use such low-end devices alone, you may encounter some manageability issues.
Workgroup level fiber switch
Fiber switches provide the ability to cascade many switches into a large-scale fabric. By connecting one or more ports of two switches, all ports connected to the switch can see the unique image of the network. Any node on the fabric can communicate with other nodes.
Essentially, through a cascade switch, it is possible to build a large, virtual, switch with a distributed advantage, and it can span a very large distance. Fabrics that are built by multiple switches look like a fabric that is made up of a separate switch. All ports on the switch can view and access all other ports on the fabric as if they were a local switch. Unified name servers and management services allow you to view and modify all fabric information through a separate interface.
An important factor in creating a distributed fabric is to get the bandwidth of the connection between the switches. The effective rate between any two ports is affected by the effective bandwidth of the connection between the switches and may require the use of connections between multiple switches to maintain the necessary bandwidth. Workgroup Fiber Channel switches are numerous and more versatile.
Users can use workgroup switches for multiple ways, but the most widely used areas are small SANs. This type of switch can be connected by interconnection between the switches to provide more ports. Interconnections between switches can be created on any port on a Fiber Channel switch. However, if you plan to use multiple manufacturers of products, be sure to ensure that equipment can be interoperable.
Core level fiber switch
Core-level switches (also known as directors) are generally located in the center of a large SAN, so that several edge switches are connected to each other to form a SAN network with hundreds of ports. Core switches can also be used as a separate switch or edge switch, but it enhances functionality and internal structure to make it work better in a core storage environment. Other features of the core switch include support for protocols other than fiber (like InfiniBand), support for 2Gbps Fiber Channel, advanced fiber services (eg, security, trunk and frame filtering, etc.).
Core-level fiber switches typically provide many ports, from 64 to 128 ports to more. It uses a very wide internal connection to route data frames with maximum bandwidth. The purpose of these switches is to build a larger coverage network and provide greater bandwidth, which is designed to route frame signals with the shortest possible delay between multiple ports at as fast a speed as possible.
In addition, the core fiber switch is often based on "blade" hot-swappable circuit board: as long as the switchboard inserted in the cabinet can add the necessary new features, you can also make online maintenance, you can also do online phased press Need to expand. Many core-level switches do not support arbitration loops or other direct-connected loop devices, which only focus on core switching capabilities.
Because the usability is the most important in the whole environment, people are willing to spend more money to buy the redundancy, all the components of the high redundant switch are redundant, completely remove the single point of failure, and to ensure a very long Normal running time. These expenses on redundancy are typically spent on high availability backplanes, power supplies, redundant circuits, and maintenance availability. This type of switch built a lot of logic circuit, used to deal with the internal hardware failure of the switch.
In addition to redundancy, core-class fiber switches support uninterrupted service-oriented software upgrades, eliminating the need for system maintenance at the time of the upgrade. The alternate path is a redundancy level on the network that can be configured with a resilient dual fabric that completely eliminates a single point of failure and avoids the failure of the network due to software or hardware errors, fires, natural disasters, or operational errors To the serious consequences.
Core-level switches provide the highest level of reliability and port density. In a data center with a large number of Fiber Channel infrastructures, these products are almost invariable, centralized storage switches. So, for most high-availability networks, you should choose a dual-channel network built by a core fiber switch.