|Hardware Redundancy |
One way to ensure uptime is to employ load sharing, redundant power supplies in switching, routing and media converter chassis. Most manufacturers offer these types of options for their chassis based solutions. Another way to ensure power redundancy is to use a backup UPS unit. Should the main power fail, your systems would stay powered via battery packs in the UPS unit.
Hot Swap Modules
Many GBIC and switching or routing blades are designed to be hot swappable meaning they can be replaced on the fly while the main system/chassis is still powered up. This saves time when a component on a blade or module fails.
|Back planes |
Another redundant hardware feature on some chassis is the passive data back plane. These back planes, which carry all the systems communications data, do not have any active components that can fail so by design they eliminate a single point of failure.
Some manufacturers offer standby fail over ports for physical redundancy. This scenario uses a combination
of hardware and software in
the switch or router to use the secondary port when a special packet is not seen in a specified time frame on the primary port.
There are several redundant standards based and proprietary ring designs, which provide backup paths for data to travel on in the event of a broken or pulled cable. Some people choose to design their whole network around a redundant ring technology such as FDDI or Token Ring. Both of these technologies use a physical ring architecture with redundant ports and a token passing design which will route data the opposite way if a certain packet or token is not seen in a certain amount of time due to a broken or pulled cable or failure of the attached equipment. Companies such as Hirschmann and Garrettcom utilize homegrown, proprietary, redundant ring schemes in their equipment.
The most common and widely used is the redundant connection protocol called Spanning Tree Protocol which is used in Ethernet networks. Spanning Tree automatically assigns a standby port which can be used to re-route traffic in the event one of the other inter switch links fails or is broken. Several more recent variations of Spanning Tree Protocol have been created which support faster fail over times such as Rapid Spanning Tree Protocol (RSTP) and Spanning tree per VLAN or subnet PVST (Per VLAN Spanning Tree) which provides small groups of Spanning trees in a network thereby increasing the total number of fail over links.
|By utilizing a combination of hardware and software redundancy in your network you can almost guarantee 100% resiliency and uptime. |