Tuesday, September 15, 2009

Metro Ethernet is Gaining Speed

I remember in 2003-2004, MPLS was the big buzzword going around IT shops. Almost as controversial as VOIP, it created a lot of debate between Network Architects and WAN Engineers. Architects liked it because it was any-to-any (no PVC’s), supported Quality of Service (QoS) for voice/video, and depending on carrier, last-mile agnostic (could be Frame Relay, PPP, ATM, Ethernet, etc.). Engineers liked it for the above reasons too. However, they felt they lost some control and visibility of the network once it entered the carrier core IP Cloud.

Of course the main benefit was lower cost so these networks got hugely popular and by 2008, 80% of enterprise customers had some version of MPLS with their network. MPLS has actually become the de facto WAN standard.

I do see another WAN or MAN product picking up a lot of steam especially for regional companies – Metropolitan Ethernet. The three big benefits are:

1. Scalable bandwidth. I can grow from 10 Mg to 100 Mg – or even a Gig without having to upgrade equipment or deploying a lot of resources to make this happen.
2. Cost. A DS3 (45 Mg) of MPLS service cost in the $5500 to $6000 range. I have seen 100 Mg Metro E product for $1800. WOW! You can get twice the bandwidth for one third the price.
3. Local diversity. There are regional players who have glass in the ground that are a legitimate solution and can offer a great replacement or enhancement to the Local Exchange Carrier (LEC). We are actually seeing a lot of growth in this area using a Carrier A and Carrier B approach where Carrier A is the LEC and Carrier B is an alternative company (cable or power). Hopefully, the routes/paths are diverse. This avoids a lot of special construction cost.

Metropolitan Ethernet does have its limitations. For one thing, as the name implies, it is “Metro” so if you need to connect Miami and San Francisco it might be the right solution. But it can be used as a transport method to a core network (MPLS?). Also some Metro E players can extend outside of the LATA gaining great cost efficiencies.

Another limitation is QoS. Metropolitan Ethernet delivers Class of Service (COS), not QoS. With CoS, the network administrator must ensure the network will not be over-committed with high-priority traffic. If you have a Gig link and allow 100 Mbps of priority traffic on the link, CoS should work.

LME Consulting has experience in designing, provisioning, and supporting WAN/MAN products from a variety of carriers. Call us at 916-601-1961 or e-mail us at rich@lmeconsulting.net to help you design your next network.

I am Sold on Hosted VOIP

We have been evaluating Hosted VOIP products for months for resell purposes. I had been skeptical of hosted systems for a few reasons; latency, cost structure, and overall reliability. However, market demands require us to have a product so we chose VOIPStreet because of their price structure and call quality.

We have been tickled with how well the product works. It allows business to do things that literally transforms the business. Many of the constraints are gone, multiple locations can look like one and customer service calls get handled properly. The call quality is not as good as the PSTN but it is definitely cell-phone quality.

The cost is what is so impressive; $100 for ten extensions (with ten Voice Mail boxes), two lines, free local, free long distance, and most of the features you would expect in a state-of-the-art PBX. Any SIP compliant phone will work, however, a good percentage of our customer’s forward extensions to cell phones. Set-up time blew me away. We had a small system built within 20 minutes.

I see all kinds of opportunities for the following customer scenarios.

1. Any customer with multiple locations.
2. New business with uncertain growth expectations.
3. Customers that need a temporary system.

LME Consulting is a certified VOIPStreet partner. Call us at 916-756-4258 or email us at rich@lmeconsulting.com and we will help you design, plan, and install your communications system.