September 2010 NetEqualizer News Ireland Tech Seminar Coming Soon; Starbucks Gift Card Offer Announced Greetings! Enjoy another issue of the NetEqualizer Newsletter. This month, we announce more details about October’s Ireland Tech Seminar and kickoff our most recent Starbucks gift card offer. As always, feel free to pass this along to others who might be interested in NetEqualizer news. In this issue: October Ireland Tech Seminar Coming Soon! What Is The NetEqualizer Doing For You? NetEqualizer Trivia Best Of The Blog
October Ireland Tech Seminar Coming Soon!
		We have now opened registration for our Fall Technical Seminar in Dublin, Ireland, presented by our authorized reseller Ai Bridges. For your convenience, we’re offering the Seminar on two dates: Monday, October 4th, and Tuesday, October 5th. Sign up for whichever date works best for you! This seminar is not to be missed…Our CTO, Art Reisman, will be there in person to discuss the NetEqualizer technology and to answer your questions. Here are the details: When: Monday, October 4, 2010, OR Tuesday, October 5, 2010 9:30 a.m. to 3:15 p.m. (same time both days) Lunch will be provided Where: Dublin’s Burlington Hotel (click here for map) CLICK HERE TO REGISTER We look forward to seeing you this October in Dublin, Ireland!
What Is The NetEqualizer Doing For You?
		At APconnections, we like to keep up with how the NetEqualizer is being used in different environments around the world. Over the past several years, a major part of how we spread the word about the NetEqualizer has been through the experiences of our customers.So, if you’re currently using the NetEqualizer and want to share your experience with the technology, send us a quick one-paragraph write-up on what the NetEqualizer is doing for you. As a “thank you” for your time, we’ll send you a $20 Starbucks gift card. Testimonials should be sent to admin@apconnections.net. We look forward to hearing what you have to say! (Limited supply of gift cards available. Offer good while supplies last. One gift card per customer.)
NetEqualizer Trivia
		While many of our customers know us pretty well, here are a few NetEqualizer facts that are sure to leave some of you stumped. The correct answers can be found at the end of the newsletter. 1. Where is NetEqualizer manufactured? Beijing Austin, Texas Santa Clara, California Hong Kong 2. What country outside of North America has the most NetEqualizer systems in use? France Ireland China India 3. The NetEqualizer can be found in which of these locations? Faroe Islands Malta Iceland Seychelles 4. How many end users have their Internet service go through a NetEqualizer on an average day? 10 million 4 million 2 million 500,000
Best Of The Blog
		Top Five Causes For Disruption Of Internet Service Editor’s Note: We recently took a poll from our customer base consisting of thousands of NetEqualizer users. What follows are the top five most common causes for disruption of Internet connectivity. 1) Congestion Congestion is the most common cause for short Internet outages. In general, a congestion outage is characterized by 10 seconds of uptime followed by approximately 30 seconds of chaos. During the chaotic episode, the circuit gridlocks to the point where you can’t load a Web page. Just when you think the problem has cleared, it comes back. The cyclical nature of a congestion outage is due to the way browsers and humans retry on failed connections. During busy times usage surges and then backs off, but the relief is temporary. Congestion related outages are especially acute at public libraries, hotels, residence halls and educational institutions. To keep reading, click here.
		Contact Information email: admin@apconnections.net phone: 303-997-1300 web: http://www.netequalizer.com Join our mailing list

Trivia Answers

1. Santa Clara, California – Since switching to ASA Computers for our semi-custom manufactured servers in 2005, NetEqualizer has been completely U.S.-based and manufactured. ASA, who provides full solutions including ‘ISP packages’, servers for ‘.com’ companies, computing environments for schools and research organizations, RAID and NAS set-ups, network installation, help-desk support, and much more, also makes equipment for the likes of Cisco, Microsoft and many Fortune 500 companies.
2. Ireland (the UK is second)
3. All of the above
4. 4 million users

Have you ever logged into your wireless laptop at a library or hotel lobby or airport?

Have you ever visited and used WiFi in a small-town coffee shop?

Do you take classes at a local university?

What got us thinking on this subject was the flurry of articles on net neutrality — a hot-button issue in the media these days. With each story, the reporters usually rush to get quotes and statements from all the usual suspects — Verizon, Google, Comcast, Time Warner, etc. It’s as if these providers ARE the Internet. However, in this article, we’ll show there is a significant loose conglomerate of smaller providers that, taken together, create a much larger entity than any of these traditional players.

These smaller organizations buy bulk bandwidth from tier-1 providers such as Level 3 and then redistribute it to their customers. In other words, they are your ISP. To give you a rough idea on just how large this segment is, we have worked up some numbers with conservative estimates.

There are roughly 121,000 libraries in the US. Some are very large with thousands of patrons per day and some are very small with perhaps just a handful of daily visitors. We estimate that half provide some form of wireless Internet service, and of those, they would average 300 unique users per month. That gives us approximately 18 million patrons using the Internet in libraries per year.

There are approximately 15 million students attending higher education institutions, with K-through-12 schools making up another 72 million students. If all the university students, and perhaps half of the K-through-12 students use the Internet at their schools, that gives us another 45 million users.

In 2004, half the hotels in the U.S. had broadband service.  It would be safe to assume that this numbers is over 90 percent in 2010. There are approximately 130,000 hotels listed in the US. With an average occupancy per night of 30 guests per hotel (very conservative), we can easily conclude that 100 million people use the Internet from U.S. hotels over the course of a year.

Lastly there are 10,000 small regional ISPs and cable companies serving smaller and rural customers. These companies average about 1,000 customers, covering another 10 million people.

Yes, some of these users are being double counted as many obviously have multiple sources to the Internet, but the point is, with conservative estimates, we were able to easily estimate 100 million users through these alternate channels, making this segment much larger than any single provider.

Therefore, when discussing the issue of net neutrality, or any regulation or privacy debate concerning the Internet, one should look beyond just the big-name providers. There’s a good chance you’ll find your own online experience regularly extends beyond these high-profile ISPs.

NetEqualizer bandwidth controllers are used in hotels, libraries, schools, WiFi hotspots and businesses around the world and have aided in the Internet experience of over 100 million users since 2003.

Editor’s Note: Looks like the metered bandwidth is back in the news. We first addressed this subject back in June of 2008. Below you’ll find a few articles on the topic, followed by our commentary.

The Party’s Over: The end of The Bandwidth Buffet

By Brian Santo, Editor-in-Chief, CedMagazine.com – September 01, 2010

As the consumption of video on broadband accelerates, moving to consumption billing is the only option.

Arguments over consumption billing and network neutrality flared up again this summer. The associative connector of the two issues is their technical underpinning: Consumption billing is based on the ability to measure, meter and/or monitor bits as they flow by. The problem is that those abilities are what worry some advocates of one version of network neutrality.

The summer season began with AT&T stirring things up with an announcement that it was moving toward adopting consumption billing for wireless broadband.

To keeping, click here.

Internet Providers Want To Meter Usage: Customers Who Like To Stream Movies, TV Shows May Get Hit With Extra Fees

By Dan Richman, msnbc.com contributor – September 15, 2009

If Internet service providers’ current experiments succeed, subscribers may end up paying for high-speed Internet based on how much material they download. Trials with such metered access, rather than the traditional monthly flat fee for unlimited connection time, offer enough bandwidth that they won’t affect many consumers — yet…

For the full article, click here.

Here is our original commentary on the subject:

The recent announcement that Time Warner Cable Internet plans to experiment with a quota-based bandwidth system has sparked lively debates throughout cyberspace. Although the metering will only be done in a limited market for now, it stands as an indication of the direction ISPs may be heading in the future. Bell Canada is also doing a metered bandwidth approach, in Canada much of the last mile for Bell is handled by resellers and they are not happy with this approach.

Over the past several years, we have seen firsthand the pros and cons of bandwidth metering. Ultimately, invoking a quota-based system does achieve the desired effect of getting customers to back off on their usage — especially the aggressive Internet users who take up a large amount of the bandwidth on a network.

However, this outcome doesn’t always develop smoothly as downsides exist for both the ISP and the consumer. From the Internet provider perspective, a quota-based system can put an ISP at a competitive disadvantage when marketing against the competition. Consumers will obviously choose unlimited bandwidth if given a choice at the same price. As the Time Warner article states, most providers already monitor your bandwidth utilization and will secretly kick you offline when some magic level of bandwidth usage has been reached.

To date, it has not been a good idea to flaunt this policy and many ISPs do their best to keep it under the radar. In addition, enforcing and demonstrating a quota-based system to customers will add overhead costs and also create more customer calls and complaints. It will require more sophistication in billing and the ability for customers to view their accounts in real time. Some consumers will demand this, and rightly so.

Therefore, a quota-based system is not simply a quick fix in response to increased bandwidth usage. Considering these negative repercussions, you may wonder what motivates ISPs to put such a system in place. As you may have guessed, it ultimately comes down to the bottom line.

ISPs are often getting charged or incurring cost overruns on total amount of bytes transferred. They are many times resellers of bandwidth themselves and may be getting charged by the byte and, by metering and a quota-based system, are just passing this cost along to the customers. In this case, on face value, quotas allow a provider to adopt a model where they don’t have to worry about cost overruns based on their total usage. They essentially hand this problem to their subscribers.

A second common motivation is that ISPs are simply trying to keep their own peak utilization down and avoid purchasing extra bandwidth to meet the sporadic increases in demand. This is much like power companies that don’t want to incur the expense of new power plants to just meet the demands during peak usage times.

Quotas in this case do have the desired effect of lowering peak usage, but there are other ways to solve the problem without passing the burden of byte counting on to the consumer. For example, behavior-based and fairness reallocation has proven to solve this issue without the downsides of quotas.

A final motivation for the provider is that a quota system will take some of the heat off of their backs from the FCC. According to other articles we have seen, ISPs have discreetly, if not secretly, been toying with bandwidth, redirecting it based on type and such. So, now, just coming clean and charging for what consumers use may be a step in the right direction – at least where policy disclosure is concerned.

For the consumer, this increased candor from ISPs is the only real advantage of a quota-based system. Rather than being misled and having providers play all sorts of bandwidth tricks, quotas at least put customers in the know. Although, the complexity and hassle of monitoring one’s own bandwidth usage on a monthly basis, similar to cell phone minutes, is something most consumers most likely don’t want to deal with.

Personally, I’m on the fence in regard to this issue. Just like believing in Santa Claus, I liked the illusion of unlimited bandwidth, but now, as quota-based systems emerge, I may be faced with reality. It will be interesting to see how the Time Warner experiment pans out.

Related:  Blog dedicated to stamping out usage based billing in Canada.

Editor’s final note: We are also seeing renewed interest in Quota based system, we have completely revamped our NetEQualizer Quota interface this spring to meet rising demand.

Created by APconnections, the NetEqualizer is a plug-and-play bandwidth control and WAN/Internet optimization appliance that is flexible and scalable. When the network is congested, NetEqualizer’s unique “behavior shaping” technology dynamically and automatically gives priority to latency sensitive applications, such as VoIP and email. Click here for a full price list.

After inquiring a bit deeper, I also found out this customer was using extensive VPNs to remote offices over DSL internet circuits. His WAN traffic from the remote offices was sharing links with regular Internet data traffic, and all of it was traversing the public Internet. Then it hit me – he did not realize his QoS mechanisms were useless outside of his internal network.

Where there is one customer with confusion there are usually others. Hence, I’ve put together a quick fact sheet on QoS over an internet link. Below, you’ll find five quick facts that should help clarify QoS and answer the primary question of it’s possible over the Internet.

Fact #1

If your QoS mechanism involves modifying packets with special instructions (ToS bits) on how it should be treated, it will only work on links where you control both ends of the circuit and everything in between.

Fact #2

Most Internet congestion is caused by incoming traffic. For data originating at your facility, you can certainly have your local router give priority to it on its way out, but you can’t set QoS bits on traffic coming into your network (We assume  from a third party). Regulating outgoing traffic with ToS  bits will not have any effect on incoming traffic.

Fact #3

Your public Internet provider will not treat ToS bits with any form of priority (The exception would be a contracted MPLS type network). Yes, they could, but if they did then everybody would game the system to get an advantage and they would not have much meaning anyway.

Fact #4

The next two facts address our initial question — Is QoS over the Internet possible? The answer is, yes, QoS on an Internet link is possible. We have spent the better part of seven years practicing this art form and it is not rocket science, but it does require a philosophical shift in thinking to get your arms around.

We call it “equalizing,” or behavior-based shaping, and it involves monitoring incoming and outgoing streams on your Internet link.  Priority or QoS is nothing more than favoring one stream’s packets over another stream’s. You can accomplish priority QoS on incoming streams by queuing (slowing down) one stream over another without relying on ToS bits.

Fact #5

Surprisingly, behavior-based methods such as those used by our NetEqualizer do provide a level QoS for VoIP on the public Internet. Although you can’t tell the Internet to send your VoIP packets faster, most people don’t realize the problem with congested VoIP is due to the fact that their VoIP packets are getting crowded out by large downloads. Often, the offending downloads are initiated by their own employees or users. A good behavior-based shaper will be able to favor VoIP streams over less essential data streams without any reliance on the sending party adhering to a QoS scheme.

For more information, check out Using NetEqualizer To Ensure Clean Clear VOIP.

1) Congestion

Congestion is the most common cause for short Internet outages.  In general, a congestion outage is characterized by 10 seconds of uptime followed by approximately 30 seconds of chaos. During the chaotic episode, the circuit gridlocks to the point where you can’t load a Web page.  Just when you think the problem has cleared, it comes back. The cyclical nature of a congestion outage is due to the way browsers and humans retry on failed connections. During busy times usage surges and then backs off, but the relief is temporary.

Congestion related outages are especially acute at public libraries, hotels, residence halls and educational institutions.

2) Failed Link to Provider

If you have a business-critical Internet link, it’s a good idea to source service from multiple providers. Between construction work, thunderstorms, wind, and power problems, anything can happen to your link at almost any time. These types of outages are much more likely than internal equipment failures.

3) Service Provider Internet Speed Fluctuates

Not all DS3 lines are the same. We have seen many occasions where customers are just not getting their contracted rate 24/7 as promised.

4) Equipment Failure

Power surges are the most common cause for frying routers and switches. Therefore, make sure everything has surge and UPS protection. After power surges, the next most common failure is lockup from feature-overloaded equipment. Considering this, keep your configurations  as simple as possible.

5) Operator Error Mistake

Duplicating IP addresses, plugging wires into the wrong jack, and setting bad firewall rules are the leading operator errors reported.

Editor’s Note: If you commonly encounter issues that aren’t discussed here, feel free to fill us in in the comments section. For more information on disruption of service, take a look at the following articles: http://www.onecomputerguy.com/networking/troubleshoot_internet.htm.

http://technologizer.com/2008/08/11/eight-great-internet-outages/

Yes, I believe that most people who work for Google and Verizon, executives included, believe in an open Neutral Internet.  And yet, from experience, when push comes to shove, and profits are flat or dropping, the idea of leveraging your assets will be on the table.  And what better way to leverage your assets than restrict competition to your captive audience. Walling off a captive audience to selected content will always be enticing to any service provider looking for low hanging fruit.  Morals can easily be compromised or rationalized in the face of losing your house, and it only takes one over zealous leader to start a provider down the slope.

The checks and balances so far, in this case, are the consumers who have voiced outright disgust with anybody who dare toy with the idea of  preferential  treatment of Internet traffic for economic benefit.

For now this concept will have to wait, but it will be revisited again and hopefully consumers will rise up in disgust.  It would be naive to think that today’s statement by Verizon and Google would be  binding beyond the political moment.

Editor’s note: Art Reisman is the CTO of APconnections. APconnections designs and manufactures the popular NetEqualizer bandwidth shaper.

Editors Note: The official term for one public IP address mapped to multiple private IP addresses is PAT. However, most IP people use the terms interchangeably.

I was doing some internal research on natural address translation (NAT) this past week, and as I looked for reliable sources, I became a bit frustrated with the information available. Yes, the information is out there and the Wikipedia article has some nice charts with all the details. But, if you’re looking  for the rational reasons behind NAT, you might want to shoot yourself in the head by the time you read through all of the information and find what you’re looking for.

To preserve your sanity, as well as answer some key questions quickly, I’ve put together the following Q&A detailing some key points when it comes to NAT. We’ll start with the basics and go from there.

What is NAT?

In order to allow multiple users to share a single IP address, modern routers utilize NAT to find unused port numbers and map them to a set of local private IP addresses. So, for example, let’s say your Internet provider gives you a single IP address for your household. It could be something like 98.245.90.60, which is a public IP address owned by Comast.

All of the computers in your house must share the single IP address that Comcast provides. So, your local router — the Linksys wireless router you bought for $79 — will use NAT to tag traffic with port numbers and then create some additional IP addresses right where your house connects to the Internet.

Let’s say you contacted the Microsoft website to download the latest service pack. When Microsoft sends you the download, it’s going to send it to 98.245.90.60:5001. “5001″ is the port number established for the FTP transfer and 98.245.90.60 is the Comcast-owned Internet address for your entire house. Using NAT, your router will then interpret the port number and change the IP address to a unique internal address (like 192.168.1.103:8700, for example) before it gets to your computer.

Why do we need NAT?

NAT is useful because home users often have more than one computer in their household and yet only have a single IP address from their provider. Since every computer that talks on the Internet requires an IP address, it would not be possible to have more than one computer in your house without NAT.

How does NAT map a single IP address to multiple computers without things like Web browsing getting mixed up?

First, here’s some background on the difference between a base IP address and a port number. Internet addresses have two parts: an IP address, such as 98.243.90.60, and a port number. The IP address is used to route data across the Internet and the port is used by the receiving device — your computer — to determine what service to provide. For example, port 80 is the default port address for Web browsing.

Before the invention of NAT, Internet routers mostly ignored the port part of the address as they did not need it to move IP packets across the Internet. When describing the function of a port number, I like to use the analogy of a large dormitory with individual room numbers for the people living there. The postal service ignores the room numbers as their service ends at the address of the dormitory. They do not sort the mail by room number. For internet routers, port numbers are like room numbers. They deliver the packet to the end user’s computer and the port number is then interpreted.

The range of possible port numbers are in the tens of thousands, which is more than enough interpreting services by a user’s computer.  Think of a dorm with 1,000 residents in which they would only need 1,000 numbers for mailboxes, but still had 1,000,000 reserved.

What happens if there are no free ports to do the translation?

On small home networks this is not likely to happen, but you can get conflicts if, for example, you try to use NAT on a network with tens of thousands  of users. The total number of unique ports available is 65,000 and most users will require more than one port at a time.

Does NAT slow down my Internet connection?

Not enough for you to notice.

Why does my provider only allocate one IP address for my residence?

Even though there are about 4,000,000,000 (four billion) possible Internet addresses, the actual addresses are given out in large blocks, and once given out, they are hard to get back. So, and this is purely an example, let’s say a large company was given a class B set of addresses (which used to be common in the early days). They would have 64,000 addresses in their control. Hence, even with 4,000,000,000 possible addresses, they are in short supply, and your provider cannot afford to give them out more than one at a time.

Can I have more than one IP address?

Yes, but you would likely need a business class Internet service, which is generally quite a bit more expensive than residential-type service.

When will the world run out of IP addresses?

Some say we already have and there is a big push to go to a new standard called IPV6. However, we don’t think that will ever happen.

Editors Note: The official term for one public IP address mapped to multiple private IP addresses is PAT. However, most IP people use the terms interchangeably.

Are you seeing a drop off in layer 7 bandwidth shapers in the marketplace?

In the early stages of the Internet, up until the early 2000s, the application signatures were not that complex and they were fairly easy to classify. Plus the cost of bandwidth was in some cases 10 times more expensive than 2010 prices. These two factors made the layer 7 solution a cost-effective idea. But over time, as bandwidth costs dropped, speeds got faster and the hardware and processing power in the layer 7 shapers actually rose. So, now in 2010 with much cheaper bandwidth, the layer 7 shaper market is less effective and more expensive. IT people still like the idea, but slowly over time price and performance is winning out. I don’t think the idea of a layer 7 shaper will ever go away because there are always new IT people coming into the market and they go through the same learning curve. There are also many WAN type installations that combine layer 7 with compression for an effective boost in throughput. But, even the business ROI for those installations is losing some luster as bandwidth costs drop.

So, how is the NetEqualizer doing in this tight market where bandwidth costs are dropping? Are customers just opting to toss their NetEqualizer in favor of adding more bandwidth?

There are some that do not need shaping at all, but then there are many customers that are moving from $50,000 solutions to our $10,000 solution as they add more bandwidth. At the lower price points, bandwidth shapers still make sense with respect to ROI. Even with lower bandwidth costs  users will almost always clog the network with new more aggressive applications. You still need a way to gracefully stop them from consuming everything, and the NetEqualizer at our price point is a much more attractive solution.

Editor: Tell us how you got started in the bandwidth control business?

Back in 2002, after starting up a small ISP, my partners and I were looking for a tool that we could plug in and take care of the resource contention without spending too much time on it. At the time, we had a T1 to share among about 100 residential users and it was costing us $1200 per month, so we had to do something.

Editor: So what did you come up with?

I consulted with my friends at Cisco on what they had. Quite a few of my peers from Bell Labs had migrated to Cisco on the coat tails of Kevin Kennedy, who was also from Bell Labs. After consulting with them and confirming there was nothing exactly turnkey at Cisco, we built the Linux Bandwidth Arbitrator (LBA) for ourselves.

How was the Linux Bandwidth Arbitrator distributed and what was the industry response?

We put out an early version for download on a site called Freshmeat. Most of the popular stuff on that site are home-user based utilities and tools for Linux. Given that the LBA was not really a consumer tool, it rose like a rocket on that site. We were getting thousands of downloads a month, and about 10 percent of those were installing it someplace.

What did you learn from the LBA project?

We eventually bundled layer 7 shaping into the LBA. At the time that was the biggest request for a feature. We loosely partnered with the Layer 7 project and a group at the Computer Science Department at the University of Colorado to perfect our layer 7 patterns and filter. Myself and some of the other engineers soon realized that layer 7 filtering, although cool and cutting edge, was a losing game with respect to time spent and costs. It was not impossible but in reality it was akin to trying to conquer all software viruses and only getting half of them. The viruses that remain will multiply and take over because they are the ones running loose. At the same time we were doing layer 7, the core idea of Equalizing,  the way we did fairness allocation on the LBA, was s getting rave reviews.

What did you do next ?

We bundled the LBA into a CD for install and put a fledgling GUI interface on it. Many of the commercial users were happy to pay for the convenience, and from there we started catering to the commercial market and now here we are with modern version of the NetEqualizer.

How do you perceive the layer 7 market going forward?

Customers will always want layer 7 filtering. It is the first thing they think of from the CIO on down. It appeals almost instinctively to people. The ability to choose traffic  by type of application and then prioritize it by type is quite appealing. It is as natural as ordering from a restaurant menu.

So why did you get away from layer 7 support in the NetEqualizer back in 2007?

When trying to contain an open Internet connection it does not work very well. The costs to implement were going up and up. The final straw was when encrypted p2p hit the cloud. Encrypted p2p cannot be specifically classified. It essentially tunnels through $50,000 investments in layer 7 shapers, rendering them impotent. Just because you can easily sell a technology does not make it right.

We are here for the long haul to educate customers. Most of our NetEqualizers stay in service as originally intended for years without licensing upgrades. Most expensive layer 7 shapers are mothballed after about 12 months are just scaled back to do simple reporting. Most products are driven by channel sales and the channel does not like to work very hard to educate customers with alternativetive technology. They (the channel) are interested in margins just as a bank likes to collect fees to increase profit. We, on the other hand, sell for the long haul on value and not just what we can turn quickly to customers because customers like what they see at first glance.

Are you seeing a drop off in layer 7 bandwidth shapers in the marketplace?

In the early stages of the Internet up until the early 2000s, the application signatures were not that complex and they were fairly easy to classify. Plus the cost of bandwidth was in some cases 10 times more expensive than 2010 prices. These two factors made the layer 7 solution a cost-effective idea. But over time, as bandwidth costs dropped, speeds got faster and the hardware and processing power in the layer 7 shapers actually rose. So, now in 2010 with much cheaper bandwidth, the layer 7 shaper market is less effective and more expensive. IT people still like the idea, but slowly over time price and performance is winning out. I don’t think the idea of a layer 7 shaper will ever go away because there are always new IT people coming into the market and they go through the same learning curve. There are also many WAN type installations that combine layer 7 with compression for an effective boost in throughput. But, even the business ROI for those installations is losing some luster as bandwidth costs drop.

So, how is the NetEqualizer doing in this tight market where bandwidth costs are dropping? Are customers just opting to toss their NetEqualizer in favor of adding more bandwidth?

There are some that do not need shaping at all, but then there are many customers that are moving from $50,000 solutions to our $10,000 solution as they add more bandwidth. At the lower price points, bandwidth shapers still make sense with respect to ROI. Even with lower bandwidth costs  users will almost always clog the network with new more aggressive applications. You still need a way to gracefully stop them from consuming everything, and the NetEqualizer at our price point is a much more attractive solution.