How to Survive High Contention Ratios and Prevent Network Congestion

Is there a way to raise contention ratios without creating network congestion, thus allowing your network to service more users?

Yes there is.

First a little background on the terminology.

Congestion occurs when a shared network attempts to deliver more bandwidth to its users than is available. We typically think of an oversold/contended network with respect to ISPs and residential customers; but this condition also occurs within businesses, schools and any organization where more users are vying for bandwidth than is available.

 The term, contention ratio, is used in the industry as a way of determining just how oversold your network is.  A contention ratio is simply the size of an Internet trunk divided by the number of users. We normally think of Internet trunks in units of megabits. For example, 10 users sharing a one megabit trunk would have a 10-to- 1 contention ratio.
 A decade ago, a 10-to-1 contention ratio was common. Today, bandwidth is much less expensive and the average contention ratios have come down.  Unfortunately, as bandwidth costs have dropped, pressure on trunks has risen, as today’s applications require increasing amounts of bandwidth. The most common congestion symptom is  slow network response times.
Now back to our original question…
Is there a way to raise contention ratios without creating congestion, thus allowing your network to service more users?
This is where a smart bandwidth controller can help.  Back in the “old” days before encryption was king, most solutions involved classifying types of traffic, and restricting less important traffic based on customer preferences.   Classifying by type went away with encryption, which prevents traffic classifiers from seeing the specifics of what is traversing a network.  A modern bandwidth controller uses dynamic rules to restrict  traffic based on aberrant behavior.  Although this might seem less intuitive than specifically restricting traffic by type, it turns out to be just as reliable, not to mention simpler and more cost-effective to implement.
We have seen results where a customer can increase their user base by as much as 50 percent and still have decent response times for interactive  cloud applications.
To learn more, contact us, our engineering team is more than happy to go over your specific situation, to see if we can help you.

Three Myths About QoS and Your Internet Speed

Myth #1:  A QoS device will somehow make your traffic go faster across the Internet.

The Internet does not care about your local QoS device.  In fact, QoS means nothing to the Internet.  The only way your traffic can get special treatment across the Internet would be for you to buy a private dedicated link – which is really not practical for general Internet usage, as it would only be a point-to-point link.

Myth #2:  QoS will enhance the speed of your internal network.

The speed of your local internal links are a fixed rate, they always run at maximum speed.  The only way applying QoS can make something “appear” to go faster is by restricting some traffic in favor of other traffic.  I constantly get asked by our customers  if we can make important traffic get through faster, and my follow on questions are always the same.

  1. Do you have a congestion problem now?
    If not, than there is no need for any form of QoS, because your data already moving as fast as possible.
  2. If you do have congestion, what traffic do you want me to degrade so that other traffic can run without congestion?

Myth #3:  There is nothing you can do to give priority to incoming traffic on your Internet.  

Wrong! Okay, so this sounds like it may be a contradiction to Myth #1, but there is a difference in how you ask this question.   Yes, it is true that the Internet does not care about your QoS desires and will never give preferential treatment to your traffic.  But, the sending service DOES care about whether the data being transmitting is being sent at the appropriate speed for the link you get, and you can take advantage of this.

All senders of data into your network are constantly monitoring the speed at which that traffic is getting to you.  Now, if you recall the very definition of QoS is restricting one type of traffic over another.  Let’s say for example that you have a very congested Internet link with many incoming downloads.  Let’s say one download is a iOS update, and the other one is your favorite streaming Netflix movie.  By delaying the iOS update packets at the edge of your network, the sender will sense this delay, and back off on the download. The result is that there is more bandwidth left over for your favorite Netflix , and hence you have attained a higher quality of service for your Netflix over the iOS download.  How this delay is implemented is another story.

If you are interested in learning more, please feel free to contact us.

Economics of the Internet Cloud Part 1

Screen Shot 2016-04-05 at 10.07.59 AM.png

By Art Reisman

CTO, APconnections

Why is it that you need to load up all of your applications and carry them around with you on your personal computing device ?   From  I-bird Pro, to your favorite weather application, the standard operating model  assumes you purchase these things , and then  affix them to your medium of preference.

Essentially you are tethered to your personal device.

Yes there are business reasons why a company like Apple would prefer this model.   They own the hardware and they control the applications, and thus it is in their interest to keep you walled off and loyal  to your investment in Apple products.

But there is another more insidious economic restriction that forces this model upon us. And that is a lag in speed and availability of wireless bandwidth.  If you had a wireless connection to the cloud that was low-cost and offered a minimum of 300 megabits  access without restriction, you could instantly fire up any application in existence without ever pre-downloading it.  Your personal computing device would not store anything.   This is the world of the future that I referenced in my previous article , Will Cloud Computing Obsolete Your Personal Device?

The X factor in my prediction is when will we have 300 megabit wireless  bandwidth speeds across the globe without restrictions ?  The assumption is that bandwidth speed and prices will follow a similar kind of curve similar to improvements in  computing speeds, a Moore’s law for bandwidth if you will.

It will happen but the question is how fast, 10 years , 20 years 50 years?  And when it does vendors and consumers will quickly learn it is much more convenient to keep everything in the cloud.  No more apps tied to your device.  People  will own some some very cheap cloud space for all their  “stuff”,  and the  device on which it runs will become  less  and less important.

Bandwidth speed increases in wireless are running against some pretty severe headwinds which I will cover in my next article stay tuned.

Will Cloud Computing Obsolete Your Personal Device?

Screen Shot 2016-04-05 at 10.07.59 AM.png

By Art Reisman

CTO, APconnections

Twenty two years ago, all the Buzz  amongst the engineers in the AT&T Bell  labs offices,  was a technology called “thin client”.     The term “cloud” had not yet been coined yet,  but the seeds had been sowed.  We went to our project managment as we always did when we had a good idea, and as usual, being the dinosaurs that they were, they could not even grasp the concept , their brains were three sizes tooo small, and so the idea was tabled.

And then came  the Googles,  and the  Apples of the world,  the disrupters.  As bell labs reached old age , and wallowed in its death throws, I watched from afar as cloud computing took shape.

Today cloud computing is changing the face of the computer and networking world.   From my early 90’s excitement, it took over 10 agonizing years for the first cotyledons to appear above the soil. And even today,  20 years later, cloud computing is in its adolescence, the plants are essentially teenagers.

Historians probably won’t even take note of those 10 lost years. It will be footnoted as if that transition  time was instantaneous.  For those of us who waited in anticipation during  that incubation period , the time was real, it lasted over  1/4 of our professional working  lives.

Today, cloud computing is having a ripple effect on other technologies that  were  once assumed sacred. For example, customer premise networks and all the associated hardware are getting flushed down the toilet.    Businesses are simplifying their on premise networks and will continue to do so.  This is not good news for Cisco, or the desktop PC manufactures , chip makers and on down the line.

What to expect 20 years from now.   Okay here goes, I predict that the  “personal” computing devices that we know and love, might fall into decline in the next 25 years. Say goodbye to “your” IPAD or “your” iPhone.

That’s not to say you won’t have a device at your disposal for personal use, but it will only be tied to you for the time period for which you are using it.   You walk into the store , along with the shopping carts  there are  stack of computing devices, you pick one up , touch your thumb to it, and instantly it has all your data.

Imagine if  personal computing devices were so ubiquitous in society that you did not have to own one.  How freeing would that  be ?  You would not have to worry about forgetting it, or taking it through security . Where ever happened to be , in a  hotel, library, you could just grab one of the many complimentary devices stacked at the door, touch your thumb to the screen , and you are ready to go, e-mail, pictures , games all your personal settings ready to go.

Yes  you would  pay for the content and the services , through the nose most likely, but the hardware would be an irrelevant commodity.

Still skeptical ?  I’ll cover the the economics of how this transition will happen in my next post , stay tuned.

Five Requirements for QoS and Your Cloud Computing

I received a call today from one of the Largest Tier 1 providers in the world.  The salesperson on the other end was lamenting about his inability to sell cloud services to his customers.  His service offerings were hot, but the customers’ Internet connections were not.  Until his customers resolve their congestion problems, they were in a holding pattern for new cloud services.

Before I finish my story,  I promised a list of what Next Generation traffic controller can do so without further adieu, here it is.

  1. Next Generation Bandwidth controllers must be able to mitigate traffic flows originating from the Internet such that important Cloud Applications get priority.
  2. Next Generation Bandwidth controllers must NOT rely on Layer 7 DPI technology to identify traffic. (too much encryption and tunneling today for this to be viable)
  3. Next Generation Bandwidth controllers must hit a price range of $5k to $10k USD  for medium to large businesses.
  4. Next Generation Traffic controllers must not require babysitting and adjustments from the IT staff to remain effective.
  5. A Next Generation traffic controller should adopt a Heuristics-based decision model (like the one used in the NetEqualizer).

As for those businesses mentioned by the sales rep, when they moved to the cloud many of them had run into bottlenecks.  The bottlenecks were due to their iOS updates and recreational “crap” killing the cloud application traffic on their shared Internet trunk.

Their original assumption was they could use the QoS on their routers to mitigate traffic. After all, that worked great when all they had between them and their remote business logic was a nailed up MPLS network. Because it was a private corporate link, they had QoS devices on both ends of the link and no problems with recreational congestion.

Moving to the Cloud was a wake up call!  Think about it, when you go to the cloud you only control one end of the link.  This means that your router-based QoS is no longer effective, and incoming traffic will crush you if you do not do something different.

The happy ending is that we were able to help our friend at BT telecom,BT_logo by mitigating his customers’ bottlenecks. Contact us if you are interested in more details.

Six Ways to Save With Cloud Computing

I was just doing some research on the cost savings of Cloud computing, and clearly it is shaking up the IT industry.  The five points in this Webroot article, “Five Financial Benefits of Moving to the Cloud”, really hit the nail on the head.   The major points are listed below.

#1. Fully utilized hardware

#2. Lower power costs

#3. Lower people costs

#4. Zero capital costs

#5. Resilience without redundancy

Not listed in the article details was a 6th way that you save money in the cloud.  The following is from conversations I have had with a few of our customers that have moved to the Cloud.

#6.  Lower network costs

Since your business services are in the cloud, you can ditch all of those expensive MPLS links that you use to privately tie your offices to your back-end systems, and replace them with lower-cost commercial Internet links. You do not really need more bandwidth, just better bandwidth performance.  The commodity Internet links are likely good enough, but… when you move to the Cloud you will need a smart bandwidth shaper.

Your link to the Internet becomes even more critical when you go the Cloud.  But that does not mean bigger and more expensive pipes. Cloud applications are very lean and you do not need a big pipe to support them. You just need to make sure recreational traffic does not cut into your business application traffic. Here is my shameless plug: The NetEqualizer is perfectly designed to separate out the business traffic from the recreational.  Licensing is simple, and surprisingly affordable.

The NetEqualizer is Cloud-Ready.  If you are moving your business applications to the Cloud, contact us to see if we can help ease congestion for your traffic going both to and from the Cloud.

How Much Bandwidth do you Need for Cloud Services?

The good news is most cloud applications have a very small Internet footprint. The bad news is, if left unchecked, all that recreational video will suck the life out of your Internet connection before you know it.

The screen shot below is from a live snapshot depicting bandwidth utilization on a business network. Screen Shot 2016-01-27 at 12.26.49 PM

That top number, circled in red, is a YouTube video, and it is consuming about 3 megabits of bandwidth.  Directly underneath that are a couple of cloud service applications from Amazon, and they are consuming 1/10 of what the YouTube video demolishes.

Over the past few years I have analyzed quite a few customer systems, and I consistently see cloud-based business applications consuming  a small fraction of what video and software updates require.

For most businesses,  if they never allowed a video or software update to cross their network, they could easily handle all the cloud-based business applications without worry of running out of room on their trunks. Remember, video and updates use ten times what cloud applications consume. The savings in bandwidth utilization would be so great that  they could cut their contracted bandwidth allocation to a fraction of what they currently have.

Coming back to earth, I don’t think this plan is practical. We live in a video and software update driven world.

If you can’t outright block video and updates, the next best thing would be to give them a lower priority when there is contention on the line. The natural solution that most IT administrators gravitate to is to try to identify it by traffic type.  Although intuitively appealing, there are some major drawbacks with typecasting traffic on the fly.  The biggest drawback is that everything is coming across as encrypted traffic, and you really can’t expect to identify traffic once it is encrypted.

The good news is that you can reliably guess that your smaller footprint traffic is Cloud or Interactive (important), and those large 3 megabit + streams should get a lower priority (not as important).  For more on the subject of how to set your cloud priority we recommend reading: QoS and your Cloud Applications



%d bloggers like this: