Logo: TUG TORONTO USERS GROUP for Midrange Systems
TUG
e -server magazine

May 1997: Volume 12, Number 5


Communicating with Sam


WAN Cost Strategies

By Sam Johnston

From the questions submitted, the selected issue to be addressed in this issue is:

Question:

ver the last year our company has become a big user of E-mail. We have a central post office and our users currently use 1-800 dial communications to pick up their mail. The business benefits of E-mail have been great, however, my concern is that our long distance expense has been escalating. I have begun investigating using a Frame Relay network, from a carrier, for a solution to our cost problem. Justifying this solution for regional branch offices is no problem. These office are using an X.25 network for their AS/400 communications, and the Frame Relay solution would provide us with more bandwidth at a cost that is lower than what we are currently paying. The problem that I have is justifying Frame Relay for our smaller sales offices. These offices have minimal communication expense because they only communicate to the AS/400 located in the regional branch office. The regional offices are primarily local to the sales office. From the regional AS/400 the user can pass-through for access to other AS/400’s.

Having all of branches using on-line high speed communications is very attractive but I am having difficulty cost justifying the Frame Relay solution for only what I can see as a user convenience benefit. The only other issue that I can see is a benefit, is that our company is planning to implement Lotus Notes later this year, and from what I understand we will need to have a high speed WAN connection for all locations to support this environment. I believe I will have to wait until the Notes project implementation to justify implementing a Frame Relay network solution. Is there something that I can do in the interim to reduce the cost of E-mail?

Answer:

y first recommendation is that you stop viewing the Frame Relay network as a solution and see it as a service. Your business issue is cost, and a single carrier service, such as Frame Relay, is rarely a cost effective solution. It may be a case of trying to fit a square peg into a round hole. I would recommend designing a solution using services that solve your business needs versus designing your solution so you can use Frame Relay service.

There are two scenarios when a Frame Relay network service can be a cost effective single solution. First, if all locations are remote from each other. In other words, if your locations are spread across vast geography and all communication between all locations is long distance. If this is the case, then using access to a third party network (Frame Relay) can be the best solution. Remember, Frame Relay service is simply access to the providers’ Wide Area Network, where available.

The second scenario is if after you review your long distance voice communication costs and find that integrating voice with your data over the Frame Relay network would provide a cost benefit. Voice over Frame Relay technology has progressed at a rapid pace and new technological advancements from hardware vendors has made this a viable solution, from a service quality perspective. However, the design, implementation and management challenges regarding this option is best left for another edition.

From a technical design perspective a Frame Relay network can be meshed. This can be beneficial dependent upon your expected traffic management or, fault tolerant design redundancy requirements. Also, from a network management perspective a single backbone architecture can be beneficial. However, if outsourcing the network management is the objective and Frame Relay is the service that enables the network provider to provide this service then I would advise you to understand what the provider will be managing for you, and what value this provides versus the cost of using the service.

There are many high speed services available from many providers that you could use for designing your WAN Service providers. These include:

For your scenario I would first recommend that you implement a routed Frame Relay backbone with access at your regional office locations. As you mentioned, the Frame Relay solution will immediately reduce your cost while providing additional bandwidth for improved performance. As a next step I would recommend that you investigate using routed ISDN communications for the remote sales offices and use the regional office as what I call backbone hub sites. ISDN communications could be implemented between the local regional offices and the remote sales office. This connection would enable the remote offices to connect to the WAN backbone by accessing the Frame Relay backbone at the regional office.

From a cost performance perspective, ISDN communications cost is approximately $90.00 per month, per end ($180.00 per branch). An ISDN Basic Rate Interface (BRI) is comprised of two 64Kbps (B Channels) and one D Channel for signaling. The two 64Kbps channels can be bounded to provide 128kbps speed. For the regional hub sites, a benefit of Frame Relay service is its scaleability. Frame service can support from 56kbps to T1 speeds. Two other services that I could envision as possible communication solutions for the remote sales offices’ system access requirements and bandwidth needs, are: (1). the Internet, and (2). dark fiber from a Hydro utility, (depending upon their distance from the regional office).

For the specific E-mail consideration, you may want to review your current E-mail strategy. There are two things I would consider – first, your AS/400s could be used as the mail server platform. The AS/400 can be used as a POP3 mail server. A benefit of this solution is if you have a mixed desktop environment of legacy terminals and PCs, both can exchange E-mail. Second, implement local mail servers in the regional offices and distribute the mail from the central post office to these servers for the branch and remote sales office users.

In summary, your challenge is one that faces many companies today as we all move into a electronically networked world. As the electronic revolution takes hold, our networks are now strategic business assets and not just supports to our systems. The best advice that I can give to you is to design a network that will enable your business to leverage the investment for a competitive advantage. I would recommend that as a next step you work with your management and define your business needs and objectives. After this is accomplished you should design a network solution that supports these requirements. Once again, there are many services available today, and many more on the horizon, that can be integrated to meet your business requirements and, in my experience, a single service solution rarely meets all of the needs. T < G


Note: Any TUG member wishing to submit a question to Sam can e-mail or forward their typewritten material to the TUG office, or to Intesys. We would be pleased to publish your question and Sam's answer in an upcoming issue of the TUG/400 e-server magazine.