Time flies when you are having fun.
It doesnt seem that long ago that I was wrestling with a SDLC multi-drop network that spanned the Ohio and Mississippi River valleys. At the time, I thought this 42-drop network was modern. Then, Williams Telecomm introduced me to Frame Relay technology. It didnt take much convincing to make the transition from the 19.2KB multi-drop network to what was then a lightning-fast 56KB Frame Relay Network.
Frame Relay has served us well these past twenty years and continues to be the predominant wide area network technology for large corporate networks. About five years ago, broadband site-to-site VPN was the buzzword and in fact did displace a number of Frame Relay networks; however, the reliability of broadband VPN lagged behind traditional Frame Relay, and as a result, old school network managers were - and continue to be - reluctant to move away from their reliable Frame Networks.
A few years back, ATM was seen as the new technology that would displace Frame, but once again Frame Relay proved to be too reliable and economical to supplant.
Multi-Protocol Label Switching (MPLS) is the solution communication carriers are touting today. It is too soon to tell if MPLS will displace Frame Relay, however, I am betting that MPLS will indeed give Frame Relay a run for its money.
Be forewarned. The major obstacle MPLS faces is that it requires a forklift upgrade of the entire network, including local access circuits and in most cases the concentration routers and edge routers if the equipment is more than a couple years old.
From a carrier perspective MPLS is much easier to manage and maintain, so they will be pushing it aggressively and coming up with ways to make the technology financially attractive and easy to implement.
MPLS has a number of benefits including Class of Service (COS). With MPLS you can prioritize the data flow on your network and ensure the most critical data traverses the network reliably and expeditiously in congested situations. A prime candidate for priority one in a MPLS network would be VoIP, as voice traffic cannot tolerate the jitter and delay when a network becomes congested.
In a MPLS network, VoIP traffic can take precedence over any other traffic therein eliminating potential jitter and delay. Another example would be a customer facing point-of-sale (POS) application. One would not want a customer standing at the checkout counter to be inconvenienced by a transaction being held up due to a busy network. Providing a high priority class of service in a MPLS network would generally prevent this from happening.
In my opinion, the most important benefit MPLS brings to the table is the any-to-any connectivity for all points on the network. If the field office in San Diego has traffic destined for the Reno office, for example, that traffic would go directly to Reno versus going to the headquarters location and then back to Reno as it would in a traditional hub-and-spoke network.
The any-to-any traffic routing capability in and of itself creates another benefit, disaster recovery. In a MPLS network, traffic can be easily diverted to any other location on the network in the event a particular network node becomes unavailable. Of course there has to be a host available at the other selected node, but that comes with disaster recovery planning.
Lets say you have a host computer at the New York headquarters; you may decide to mirror data to a redundant system in the Chicago office. With this type of architecture, if you should lose the New York office, your transactions would automatically reroute to Chicago making the New York failure totally transparent to the remainder of the network locations.
To determine if MPLS is for you, you need to sit down and determine what type of data your network is transporting today as well as the origin and destination of the data. In addition, you need to break out the crystal ball and envision what the network will be expected to accommodate in the next few years. You should not forget VOIP, as depending on how VOIP is implemented, MPLS can prove to be a huge enabler.
If you have a traditional hub and spoke network and your assessment results determines the destination for all data is from the remote offices to the hub router site at headquarters, the value of MPLS diminishes significantly. On the other hand, if you determine there is a significant volume of traffic that originates, and is destined to, numerous points on the network then you need to take a hard look at MPLS.
The resources required and time to migrate to MPLS network as well as the ROI will no doubt be decisive factors in whether or not you choose MPLS. Over the course of the next couple of years I expect the carriers will make the migration less painful and more economical. So if you have ruled out MPLS for now, I suspect you may revisit the technology within the next couple of years.
This article was first published on EnterpriseITPlanet.com.