Bursting on an MPLS or VPLS Network

Global Ethernet VPLSA handful of carriers support bursting on their IP-VPN networks.

Bursting can provide a real cost savings benefit to customers. First, it’s a solution for customers who don’t how much bandwidth they need at a given location. Bursting lets these customers add locations to their network at the lowest level of cost commitment.

Second, it is an ideal solution for customers who know that their bandwidth needs may spike much higher than normal due to, for example, seasonal traffic peaks. Bursting lets these customers commit to the least amount of bandwidth they need for continued use and pay for only what they use in excess of that amount.

Here’s an example of how it works:

A customer commits to 100 Mbps of bandwidth at a location. The carrier sets up their IP VPN port to handle bursting. That location can now burst traffic all the way up to 1 Gbps, or whatever their local loop capacity is. For instance, you might pay for a 100M Ethernet local loop, but pay for a committed port of 20M.  This will allow you to burst to the full 100M when the need arises.

The carrier then samples traffic leaving the port throughout the month. At the end of the month, billing is calculated for the:

  • 100M local loop
  • Committed 20 Mbps port rate; plus the,
  • Sustained traffic rate in excess of the committed rate.

The carrier typically discards the top 5% of the traffic samples taken during the month. This eliminates any spurious or unusual traffic from the billing measurement.

If customers find that their sustained traffic is significantly higher than their committed data rate (in this case, 20 Mbps), they can increase their commitment and take advantage of lower prices at higher committed sustained data rates.

If you are considering changes to your Wide Area Network and would like some specialized assistance with the process, please contact us!

Why consider VPLS for your WAN

1) More flexibility and manageability with with VPLS

When it comes to rapid change and advancement, companies which can respond quickly to market shifts will  benefit from VPLS, a Virtual Private LAN Service (VPLS) solution. VPLS uses MAC addresses with Layer 2 switching as opposed to Layer 3 MPLS solutions which use IP addresses and Layer 3 routing.

The main advantage of this is that with VPLS you are in control of your own IP routing. Therefore, your IT department can be much more agile in responding to varying levels of customer demand. VPLS networks allow you to conduct rapid reconfigurations yourselves without having to contact your service provider and wait for the provider to act upon the request. Even if you do require a service provider change, the typical time to make network changes to Layer 2 VPLS networks is only a fraction of that for Layer 3 MPLS networks because the network planning process is much simpler, which could be crucial for some businesses. Another feature which aids agility is the ease of adding new sites. With a VPLS-enabled network, a new site can be added by simply changing the network router that connects the site to the VPLS network. With Layer 3 MPLS solutions, however, it is a much more complex process as all of the service provider’s routers need to be changed which typically takes 10 times as long.

2) More efficiency with VPLS

Companies with a VPLS-enabled wide area networks will be more smooth-running and thus should be able to provide a better level of service to their customers. This is down to the fact that with VPLS the company has access to its own network information so faults in a VPLS network can be isolated much faster and the IT department can trouble-shoot to fix an urgent crisis rather than having to go through a number of support engineers to get the information required from a carrier. Less network down-time means higher corporate efficiency and productivity. Another aspect of our VPLS solutions is that they offer 5 levels of Quality of Service (QoS) and allow you to define your own priority levels either through labeling your traffic or using the service aware QoS feature on the core network. This is how VPLS maximizes efficient network usage according to your business needs, so you can rest assured that mission-critical data such as CRM, ERP and SCM are allocated enough bandwidth, alongside key services such as video conferencing and telephony, even during peak usage and without costly over provisioning of network capacity.

3) Lower costs with VPLS

Companies that use VPLS solutions will find they have lower costs for a number of reasons. Firstly, VPLS enables convergence of services such as VoIP, video etc. so that all traffic can be delivered over a single Ethernet interface, eliminating multiple leased lines and resulting in economies of scale. Secondly, working with VPLS uses the same skills sets that LAN specialists have, so you would not need to provide additional training on WAN skills or hire WAN specialists. In addition, VPLS requires a lower cost CPE as it requires smaller and fewer routers than MPLS solutions.

4) Lower latencies with VPLS

As a switched, Layer 2 solution VPLS is zero-hop in the core of the network, so extremely low round-trip latencies and jitter can be achieved. For example sub 1millisecond within a metropolitan area and 67 milliseconds round-trip from London to New York. This improves the productivity of the workforce as information is available faster. It also saves retail customers using Point-of-Sale systems time dialling up to make credit/debit card payments, improving their customers’ sales experience.

Thanks to Exponential-e

 

 

 

International Capacity Price Drop to Affect Global Network Prices

When pricing a global MPLS or VPLS wide area cable network, the cost of international bandwidth has a dramatic effect on the pricing of circuits.  This is why connectivity to Asia or South America is so much more expensive that domestic circuits in the USA or circuits from the USA to Europe.  TeleGeography is a research company that compiles all this pricing data, in addition to offering some wonderful maps.

A recent wave of new submarine cable builds and upgrades to existing cable systems has brought an influx of submarine cable capacity to many historically high cost markets, including Africa, the Middle East, Southern Asia, and Latin America. Nevertheless, new data from TeleGeography show that vast regional disparities persist in both price levels and rates of decline.

New cable builds in Asia have greatly increased both supply and competition in the region, driving down prices. Median lease prices for a 10 Gbps wavelength between Los Angeles and Tokyo fell 35 percent between Q1 2011 and Q1 2012, and at a compounded rate of 33 percent between Q1 2009 and Q1 2012. Prices of 10 Gbps wavelengths between Hong Kong and Singapore fell 10 percent between Q1 2011 and Q1 2012, to $43,935 per month, and declined at a compounded 31 percent annually between Q1 2009 and Q1 2012.

Navigating the procurement of a global MPLS network is complicated unless you do this work on a daily basis, since you don’t have benchmark pricing or insight into all the global cable system.  Using MPLS-Experts to manage this process can not only save you money, but reduce the the time required to manage this process.  In many cases, we have been able to provide twice the bandwidth the customer would have obtained if they managed the process on their own.  To learn more,  visit this link or contact us.

Understanding the Federal USF Surcharge

When shopping for networks, when you compare costs to what you are paying today, look and see if your bill displays a line item for the Federal USF, the Universal Service Fund.  The surcharge changes quarterly and as of this writing, is 17.4%.  Click here to see this quarter’s FCC USF Surcharge Rate.  If  you are not paying it now, the carrier has the right to back bill you for this surcharge.  So you need to be aware when you compare costs, that your real cost will be higher than you think.

The Federal USF Surcharge should be applied to everything:

  • Sale, lease, installation and servicing of equipment.
  • Local loops
  • Network backbone

This is not a tax.  The carrier passes through their charge to you as a surcharge.

Prior to 2008, MPLS networks were not subject to the USF.  But that has changed.  And some companies that have networks contracted before 2009, may never have seen this charge on their bill…though they are liable should they be back charged by the carrier.

Despite the millions, or tens of millions, of dollars that your enterprise may have paid out in USF surcharges over the years, the surcharge has never been federally mandated on you, as a business user. The government doesn’t make you liable for AT&T or Verizon’s quarterly USF expense, no matter how much the carriers leads you to think that’s the case.  But unless your contract specifically excludes your payment of this surcharge, you are obligated to pay it, even if years have passed without seeing it on your bill.

But does it really have to be that way? Here’s where things get a little interesting. Before the FCC’s latest move, carriers had been taking different approaches to including some form of USF charge on their MPLS invoices. Those carriers that had already imposed a USF pass-through for MPLS had not necessarily applied it to all of their MPLS services. For example, some imposed it on access charges but not port charges.  But all that has changed as of 2012 and the rules are exceedingly clear.

And the truth is, there have sometimes been “seams” into which customers have been able to place themselves in order to pay less in USF charges on data services. Right before MPLS became popular, many of the nation’s big companies were buying not pure frame relay, but a service called FRASI — Frame to ATM Service Interworking. AT&T in particular tended not to charge a USF pass-through on FRASI customers, more or less officially reasoning that since one end of each connection was ATM — which, unlike frame relay, was never ruled a “basic service” — it would not impose the charge. But the larger reason was that AT&T wanted to win big frame/ATM deals vs. rivals by charging less money, and that was a good thing.  But now AT&T must access the USF, because the rules are finally clear.

So surcharges can ultimately be part of your question. If you’re not in the process of doing a deal, and you are an MPLS customer, the (harsh) reality is that your costs are probably higher than you expected, because your carrier will be paying more and will want to get it back from you. The USF surcharge itself is also going up to 17.4% of applicable revenues on April 1 and the number changes every quarter.  As of Jan 2012, there is no escaping the USF Surcharge.  Do you want to learn more, visit this web site for all the details.

 

What does OC mean? You know…OC-3, OC-12, etc.

This is a very brief post, motivated by a consulting engagement that MPLS-Experts is working on right now. This client is building a global private network to service its offices, using eight or ten collocation facilities as the Points-of-Presence. Each collo will be connected with two diverse 1Gbps Layer-1 point-to-point circuits.

So the question came up, what optical circuit do you need for 1 Gig? Not something your average client uses.

OC is short for Optical Carrier, used to specify the speed of fiber optic networks conforming to the SONET standard.

This list shows the speeds for common OC levels.
OC = Speed
OC-1 = 51.85 Mbps
OC-3 = 155.52 Mbps
OC-12 = 622.08 Mbps
OC-24 = 1.244 Gbps
OC-48 = 2.488 Gbps
OC-192 = 9.952 Gbps
OC-255 = 13.21 Gbps

We’ll need a partial OC-24 to provide 1 Gbps on each circuit.

MPLS spec introduced for cellular back-haul network service

For anyone performing cellular back-haul, there’s a new specification for handling wireless data traffic from a combination of traditional TDM networks and packet-based transport technologies as wireless operators migrate from 2G/3G to 4G and LTE services.

The Broadband Forum has just issued its “Technical Specification for MPLS in Mobile Backhaul Networks,” also known as TR-221.

TR-221 focuses on the applications of MPLS technology in a range of services that may be used to transport wireless traffic in the access and aggregation networks, including IP, TDM, ATM and Ethernet.

It defines the global requirements of MPLS technology in these networks in respect of encapsulation, signalling and routing, QoS, OAM, resiliency, security, and synchronization. It also covers expected services over the back-haul network, including voice, multimedia services, data traffic and multicast traffic, such as multimedia broadcast and multicast services (MBMS).

Adherence to these requirements will create global standards for MPLS-oriented equipment, establishing more network interoperability, speeding deployments and lowering the overall costs of the backhaul network, the Broadband Forum said.

Defining a range of reference architectures for MPLS-based mobile backhaul networks, TR-221 includes specifications for the various transport scenarios applicable to all mobile networks (2G, 3G and LTE). It also specifies the equipment requirements for the control, user and management planes to provide unified and consistent end-to-end transport services for mobile backhaul.

Robin Mersh, CEO of the Broadband Forum, said: “TR-221 is a critical part of establishing multi-vendor interoperability in converged MPLS-based backhaul networks. As mobile operators look to preserve their investment in traditional TDM and ATM networks whilst developing their 4G/LTE architectures, TR-221 will enable them to integrate new packet-based MPLS technologies into their established networks. Operators will be able to evolve their networks to be faster and more efficient to meet the increasing multimedia needs of the mobile user, whilst preserving a lower cost per bit in the backhaul network.”

Legal Tip – Avoid Auto-Renewal of Contracts With This Language

If you haven’t experienced it once in your career, you are lucky.  Carriers and other service providers often include automatic renewal language in their contracts.  After a three-year term, if you do not give notice of termination at least 30 (or 60) days before the end of your term, the contract will automatically renew for the same original term, such as three years, if that was the original term.  This language has stood up in court.

Another carrier trick is to not mention auto-renewal in the paper contract that you sign.  Instead, the contract references “terms and conditions at www.terms.carriername.com shall be incorporated into this agreement”, with the auto-renewal language displayed at this website.  Many customers do not notice this. Or, since it is not in the written agreement, in the future, it is assumed that the agreement does not auto-renew.  Wrong!

Protect yourself! Insist on including this language in your carrier contracts under the Term and Renewal section:

Upon the expiration of the Initial Term, this Agreement shall be renewed
automatically for successive Renewal Terms equivalent in duration of one (1) month
unless terminated by either Party by providing one (1) month written notice of
its intention not to renew this Agreement prior to the end of the Initial Term
or any current Renewal Term.

This language will protect you and your company from unexpected surprises several years down the road.

 

WAN using IP VPN over Internet versus MPLS – Pros and Cons

There’s a price for everything in this world, and  Internet based IP VPNs are no exception. While  IP VPNs are a cheaper alternative to any MPLS network, it doesn’t necessarily mean they’re for everyone, as customer requirements always vary. In this posting, I will explain both the Internet IP VPN advantages and disadvantages.

Let’s take a look at a few IP VPN advantages over most MPLS circuits:

  • Cheaper rates. Internet service providers provide a simple NxT1, Ethernet or Cable connection to the Internet, using the highest possible speed with. The price for internet connectivityis considerably cheaper than almost any WAN MPLS service, making it extremely attractive for companies seeking to cut telecom costs.
  • Fully configurable. WAN engineers have total control over the VPN tunnel created between sites. They are able to perform on-the-fly configuration changes to compensate for any network problems or help rectify any problem that might arise. With full access to the VPN, terminating equipment like routers and firewalls, engineers have the ability to see the condition of the internet circuit and take any action(s) deemed necessary…provided they have the staff resources and skills.
  • VPN backup included. For mission-critical sites, backup via another internet circuit is possible if your primary connection fails.  Time response for the backup line to come online is configurable by the network engineer, and there is no need to wait for the ISP to fix a line so your company can continue working.
  • Two-in-one. When configuring the site-to-site VPN, engineers can also configure remote VPN access for users traveling around the country or world, a feature most companies would have to pay additional money for to receive from their service providers.
  • Upgradable features. Perhaps one of the strongest advantages is the fact that your site-to-site VPN characteristics are strictly dependant on those that your VPN routers/firewall support. This means that as new features are introduced with the newer router operating systems (i.e., Cisco IOS), they will be available to your engineers to implement. For example, QoS pre-classification was a feature Cisco introduced in its IOS that fixed a number of QoS features for different services running over VPN tunnels. Dynamic Multiple VPN (DMVPN) was another great feature allowing scalable IPsec VPN tunnels between multiple sites. DMVPN allows every endpoint to dynamically build a VPN tunnel with any of its other peers, providing a low-cost mesh VPN solution.

If the brief list of the above  of Internet IP VPN advantages seems overwhelming , you have read a few of its disadvantages.

Here is a list of a few disadvantages of Internet IP VPNs over almost all WAN MPLS circuits:

  • Limited QoS. In order to have a fully functional QoS model, you need to have control of all equipment and paths that your VPN packets run through. In the Internet IP VPN model, QoS is effective in each site’s LAN, up until the L interface of the routers. From there on, packets enter the ISP’s network, and your ISP will clearly state that there is no QoS for such connections. Everything is based on a “best effort” delivery mechanism and you can’t argue about that. Any QoS parameters inserted in your WAN packets are, in most cases, ignored by the ISP.
  • No Class of Service Prioritization. It’s the internet, sorry.
  • Higher Packet Loss and Latency. If you use interactive applications, video, voice domestically or are connecting to locations more than 3,000 miles away, the MPLS network will outperform the IP VPN hands down.
  • Undependable voice and video. If you use voice or video over your network, the MPLS network will outperform the IP VPN, hands down with dependable and consistent performace.
  • Possible bottlenecks and low speeds. In an Internet IP VPN scenario, your company connects to the Internet, which has quite a variation of performance.  If there is heavy traffic on the Internet, chances are you might experience lower speeds during peak-hour times. Again, there is no guarantee of the performance.
  • VPN and router/firewall security. You are exposed directly to the Internet. This means that the security of your VPN and terminating equipment (routers and/or firewalls) are your responsibility. If your engineers do not take the necessary measures to secure the equipment correctly, this can lead to the exposure of your company to the Internet. This is not a topic to be taken lightly, as the damage can be devastating. It is extremely important to understand the risk involved and to have the required technical expertise to ensure the job is performed correctly. Under ideal circumstances, where the equipment is correctly configured, there is no need to worry—you’re safe.
  • Denial of service attacks. With a direct Internet connection, you are exposed to any denial of service (DoS) attack. All attempts can be successfully repelled; however, keep in mind that the traffic will have to reach your router/firewall first. This means that the heaviest damage that can be produced by a DoS attack—for a correctly configured endpoint—is to create a bottleneck on your connection and greatly reduce speeds for the duration of the attack.

If you want a rock-solid WAN with almost no packet loss and the lowest possible latency and quality, consider an MPLS network.

Global Ethernet VPN Still Limited in Many Geographies

Enterprise customers around the world are replacing legacy private line, Frame Relay, and ATM wide area networks (WAN) with Ethernet VPLS and MPLS IP VPN services. Companies’ choice of wide-area network type is shaped by a number of factors, including the applications they need to accommodate, the number of locations to be connected, the level of control the customers want to maintain over their networks, their capacity requirements, and the cost of the solution. However, data from TeleGeography’s Global Enterprise Networks Research Service suggest that the most important factor shaping an organization’s international network choice may simply be availability.

Ethernet VPN services are generally more cost effective than MPLS IP VPN services for capacity requirements above 50 Mbps, and are most appropriate for linking high-capacity headquarter sites and data centers. MPLS IP VPN
services tend to be better suited for linking large numbers of sites with more modest capacity requirements. However, dependence on Ethernet local access and the relatively slow rollout of Ethernet across MPLS PoPs means that Ethernet VPN solutions are not yet available in as many cities.

Ethernet deployments lag far behind MPLS VPN deployments, both by service provider and by geographic market. Over half of the 63 international service providers researched by TeleGeography offer MPLS VPN service in 10 or more countries, compared with less than one-third of Ethernet VPLS service providers.

The availability of IP VPN and Ethernet VPN services also differs by region. TeleGeography identified 39 IP VPN providers in Europe, 34 in Asia, 31 in the U.S. & Canada, and 19 in Africa and Latin American. Ethernet VPN services are less widely available in all of these regions, but the difference is particularly great in emerging markets. While 32 service providers offer layer 2 Ethernet VPN services in Europe, only 9 offer VPLS service in Latin America
and only 6 in Africa. While 22 service providers offer VPLS service in London, only 5 offer VPLS service in Mumbai.

This report should not limit your interest in Ethernet VPN services, but rather shape expectations on its availability.  To determine Ethernet VPN network availability for your company, contact us.

The above content provided from TeleGeography, the world’s leading independent reference source for global network infrastructure data.

MPLS, AToM and VPLS

There is often a great deal of confusion in understanding the different “flavours” of MPLS networks.  I recently read a post online by a gentleman named Mbong Ekwoge who wrote a rather clear online posting:

MPLS is the enabler of all these fancy services and applications we hear about today, such as MPLS VPNs, AToM (Any Transport over MPLS), MPLS TE (Traffic Engineering), etc.

In order to clearly understand what VPLS is, you need to understand what led to the “birth” of VPLS (Virtual Private LAN Service).  It all began with MPLS VPNs. The client had to form a peer-to-peer relationship with the Provider’s PE routers. What this means is that the provider is intricately involved with routing and forwarding the customer’s traffic and some customers did not like this idea. Also, providers had invested heavily into Layer 2 VPN techniques such as ATM, Frame Relay, etc and completely eliminating these overlay VPN techniques didn’t feel right with their financial people. Some engineers did not like the idea of having to let go of their beloved ATMs, Frame Relay PVCs for some new chap coming in.

This led Cisco and IETF to develop a solution which would let you run MPLS in the core but users will still maintain their private Layer-2 VPN service across the MPLS core of the service provider. What this means is, the provider will provide a VPN service, across MPLS, but it will be kind of a pseudowire experience. The customer still retains their highly valued privacy, the Service Provider maintains her MPLS core and should the customer be convinced, transitioning to MPLS VPNs will be like “bread and butter”.

Now this led to the introduction of AToM. AToM is the Cisco name for the Layer 2 transport service over an MPLS backbone. The customer routers interconnect with the service provider routers at Layer 2 (Ethernet, High-Level Data Link Control [HDLC], PPP, ATM, or Frame Relay). This eliminates the need for the legacy network from the service provider carrying these kinds of traffic and integrates this service into the MPLS network that already transports the MPLS VPN traffic.

AToM is an open standards-based architecture that uses the label switching architecture of MPLS and can be integrated into any network that is running MPLS. The advantage to the customer is that they do not need to change anything. Their routers that are connecting to the service provider routers can still use the same Layer 2 encapsulation type as before and do not need to run an IP routing protocol to the provider edge routers as in the MPLS VPN solution. As such, the move from the legacy network that is running ATM or Frame Relay to the network that is running AToM is completely transparent to the customer. The service provider does not need to change anything on the provider (P) routers in the core of the MPLS network. The intelligence to support AToM sits entirely on the PE routers. As such, the core and edge technologies (MPLS and AToM, respectively) are decoupled. The core label switching routers (LSRs) only switch labeled packets, whereas the edge LSRs impose and dispose of labels on the Layer 2 frames. This is similar to the MPLS VPN solution, in which the P routers switch only labeled packets and the PE routers need the intelligence to impose and dispose of labels on the IP VPN traffic from the customers.

Now how does VPLS come into the equation????

AToM is a point-to-point service and hence cannot broadcast frames.

Now some technologies such as Ethernet are broadcast in nature and take for example, the Spanning Tree Protocol (STP). These protocols operate in a broadcast nature. VPLS is the point-to-multipoint cousin of AToM.