All About IPv6 Management
Join SevOne's Scott Frymire, Director of Content Marketing, and Steve Sacchi, Product Marketing Manager at A10 Networks, as they discusses the topic of managing and deploying IPv6 technology. Questions regarding migration and integration challenges will be highlighted and answered
Good afternoon everyone than you for joining us for ensuring IT performance before, during and after your IPv6 migration. My name is Scott Frymire I'm with SevOne. I will be joined in a little bit today by my co-presenter Steve Sacchi from A10 Networks. Steve and I have been giving this presentation around the country for a while now. I know for many of you on the line today you were unable to join us when we visited your region of the country. We are very thankful that you are taking your time out of your day to join us in this webinar format today.
Before I start just a quick background on SevOne in case you haven't heard of SevOne or you are not familiar with us. Essentially what our purpose is we help organizations like yourself avoid business disruptions due to performance issues with your networks, applications and systems. We are essentially a leading significant player in the network performance and application performance management space. We provide the world’s fastest and most scalable IT monitoring and reporting solution.
Now when you talk about IPv6 migration, obviously there is an aspect of that that has to deal with managing and monitoring the deployment of this new technology and this new version of IT in your environment. I am going to talk a little bit about the SevOne product down the road here. I want to take a step back for a second. Let’s get a little bit of background about IPv6, talk about why it is a big deal for everyone in this webinar today. Then in about 30 minutes from one I’m going to hand it over to Steve. Steve will dive into a little more details, technical detail around some of the migration and the integration challenges, let’s get started.
The end is near and I'm not talking about the Mayan calendar and the supposed end of the world and what about do have about another eight days or so. I'm pretty confident we are all going to be here on December 22nd. I’m not talking about the fiscal cliff that we hear about every night on the TV on the news, no I'm talking about the end of IPv4. OK, in case you weren’t aware the world has been out of available IPv4 addresses for some time now. We have exhausted all possible IPv4 addresses. It’s a huge problem and it impacts all areas of your IT shop from your networking security, applications and so on. How did we get to this point and what are we doing about it? Well, let’s take a step back for a second. I want to start with a little background on how IP addresses are allocated just to make sure everybody has a baseline frame of reference here.
Now the IANA which is the Internet Assigned Numbers Authority they are sort of the governing body or entity that is responsible for allocating IP addresses on a global basis. What they do, is they allocate the IP addresses down to the RIRs or Regional Internet Registries. There are five of them representing different geographic regions of the world. From there the RIR allocate IP addresses down to large companies and to our internet service providers. From there the ISPs allocate them down to smaller companies and you and me. So that is how it works but back on February 3rd 2011 we exhausted all the available IPv4 addresses out there. The IANA allocated the last top level block of addresses so essentially the source it has been turned off at the top and now some of the remaining block of IP addresses are still filtering down but at the source we are out of addresses.
Now here is a map of those RIRs or Regional Internet Registries I talked about. You can see if you look over at the right hand side of the screen when you look at regions like Asia-Pacific they exhausted their IPv4 back in April 15th of 2011. Europe, Russia their addresses were exhausted back in September 14th, earlier this year. Next on the list is our area North America we are expected to run out of IPv4 addresses on June 5th 2014. I think that is about 18 months or so from now. So it is coming very fast and you might be wondering to yourselves, “Where did all these IPv4 addresses go? How did we exhaust them so quickly?” Here is a chart and I have covered up the names of the organizations have used up all the IPv4 addresses, the top organizations.
Particularly the one at the top these are represent the, the blue bars represent blocks of IP addresses that have been allocated. Each block represents 16.7 million IP addresses. The greatest offender if you will has consumed 12 blocks of IP addresses here 12 time 16.7 million. Then everybody else sort of pales in comparison. Now think for yourself for a second does anybody in their heads wants to take a guess because we do have your lines muted right now. In your head think about who do you think is the greatest usurper of IPv4 addresses? Well, it may or may not be a surprise it is the US Department of Defense. Our good old government and the DOD they are sitting on about 200 million IPv4 addresses right now. I think it is about 5% IPv4 addresses in the world, followed by Level 3 Communications, HP and then so on down the list.
Its interesting because we are actually at the point where there are companies out there when they are acquiring other companies, they are considering how many IP blocks, IPv4 blocks they have as part of the valuation of the company. It becomes a very financially enticing acquisitions sometimes if they can acquire those IP addresses with the acquisition of the company. The black market value I think nowadays is somewhere around $250 per IPv4 address. Essentially there is no more IPv4 address available as we mentioned and we need a solution. Well, enter IPv6 it is the latest revision of the internet protocol and it’s been in commercial deployment since about 2006.
Now IPv6 is a big deal because it is the only means available for the sustained on going growth of the internet. Now I’m sure all you guys are aware of IPv6, I’m going to ask you, I’m going to shoot a poll question out here. I’d like to get your feedback and tell us, what are your plans for IPv6 deployment? You should see a poll question come up and if you want to go ahead and answer whether you’re currently deploying IPv6, you are planning to deploy in the next six months next six to 12 months, 18 months or no current plans. I will leave that open for just a second. It looks like half of you there, at least half of you based on the results that we see coming in have no current plans. One or two people may be doing it in the next six months. OK so that’s good.
First off if you haven’t been planning for it yet don’t feel too bad. I've referenced our federal government and the Department of Defense. Actually the FED had imposed the deadline on themselves for IPv6 readiness. It was supposed to happen back in September of this year, a few months ago. As of today they are still only at the state of 21% IPv6 readiness when it came to all FED public facing sites and services, websites, emails, DNF. The FEDs were the largest consumer of IPv4 they haven’t even been able to meet their own mandate for when they need to get IPv6 ready. I look at the results where you guys are telling me it’s not that much of a surprise. Listen it reminds me a lot of the Y2K scare. It’s something we talk about; we have talked about for quite a while. We are getting to a point we are just about here. Like I said in 18 months we are projected to be completely out of IPv4 addresses in North America.
OK, I’m wondering how many of you have seen an IPv6 address, if you haven’t this is what it looks like. Obviously to the human eye it is a bit longer than an IPv4 address. We are dealing with eight steps of four hexadecimal characters separated by colons. We actually have a concept, I was at a trade show just a little while back we did a contest; we actually put an IPv6 address up on the screen and gave people, anybody who came by our booth they had 15 seconds to look at it. Then turn it off and id they can regurgitate it back to us we will give them an iPad. Well, you can probably guess not one person in that whole day was able to regurgitate back the IPv6 address. It is certainly not as easy to remember as an IPv4 address. Seriously how big is IPv6?
I mean visually it is certainly is, again to the human eye it is much longer and not as easily readable. When you talk about the possible combination of IPv6 addresses how big is it compared to IPv4? Let’s assume for example you see a picture of a water bottle on your screen. Let’s assume that you could fit all of the possible IPv4 address in to that one small water bottle container. The liquid in that represents all the possible IPv4 in the world, how big of a container would you need to contain all the potential of IPv6 addresses? I started sketching out the map and I thought to myself, you know what, I bet if you get something like this tanker truck with a giant tank on the back. That might be big enough to hold all the IPv6 addresses. I started doing the math and it wasn’t even close.
Try the Rose Bowl in Pasadena. If you have to fill the entire Rose Bowl from field level to the top of the stadium with water that represented all the possible IPv6 addresses, certainly that would be big enough to hold it, alright? Again not even close. You would need a container the size of planet earth to hold all the possible combinations of IPv6 addresses. I’m not talking about the oceans of the world; I’m talking about the entire volume of the planet. It’s that big and that drastic of a step up from IPv4. We are going from a 32 bit environment all the way up to 128 bit environment, about four billion addresses in IPv4 to something like 340 trillion, trillion, trillion IPv6 addresses.
Alright so this is actually the geo direct addressing becomes actually feasible in this world. Because it’s a seemingly limitless number of IP addresses that will be out there with IPv6. What is driving the need for such a large almost again seemingly infinite number of new IP addresses? Well, one of its going to be driven by world IT elements. Now let me just, there is a lot of elements on this chart here. I want to talk about a couple of points, let’s start with the red lines. Here we have the world population, this trending red line and as of today we are about seven billion in the world population. Then the bottom red line represents number of internet users. Today we are about 2½ billion internet users.
OK going back to 2010 we were significantly less, actually if we go back there I think in the year 2000 we were only about 360 million internet users in the world and today and we are up to 2½ billion. Well, if you extrapolate out the projections to the year 2020 we are expected to be a little over 5 billion internet users verses a population of let’s say, seven to eight billion. With this increase of internet users and you are considering multiple IP enabled devices for every user out there in the internet.
Whether it’s their smart phones and tablets and computers or corporate elements. When you add it all up the projection is that we are going to have nearly 20 billion corporate, 20 billion IT elements worldwide by the year 2020. OK so today we are looking at a total of about 7 billion world IT elements again up to 20 billion by the world 2020. It’s a pretty significant leap and this is what is driving the need for this many IP addresses. 2020 is not that far away when you think about it. I just think that every time I hear about the terrorist attacks of 9/11 I feel like it is almost yesterday which is not too far, that was 11 years ago.
Seven or eight years down the road is really not that far and it is coming fast. OK, here there was never enough IPv4 addresses to fulfill the current projected growth of the internet so now it is up to IPv6. We are going to change gears here just for a second. I want to ask you guys a question and you don’t have to respond as a poll but just in your own head, what do you think is the number one course of the network outages? Is it software failures, power outages, changes made by people, external viruses and servers attack or running out of IP addresses? Think about that just for a second.
Well, if you guessed C changes made by people you are correct, it is people. In fact according to research by Gardner and IDC, misconfiguration of network devices by people making unnecessary changes is the number one course of network down time for organizations. What’s even more stunning is that up to 60% of all IT problems worldwide is projected to be caused by people, people making unnecessary changes. When you think about it what does IPv6 represent? It represents change, right? It is making a change to your IP DNA to IP infrastructure.
A large part of what IPv6 represents to you guys in the IT shop is change management. You have to monitor and manage this process properly if you are to avoid the potential negative financial impact of a poor deployment. IPv6 represents quite a change it also represents something really big. Now I say big and again we talked about the number of potential IPv6 addresses out there but think of it in a different sense. Think of that chart of world IT elements that we are projected to have. We have this tidal wave coming at us of different IT element all of which are producing massive amounts of data.
With IPv6 inherently comes as need for big data network analytics. Now, I’m going to go off on a tangent here just for a second if you allow me. I mentioned big data, big data network analytics. A lot of people when they hear the term big data if you are not the IBMs of the world, you may think big data is not really for me or my organization, I’m not that large it doesn’t impact me. Consider this, big data what is it? It is just a term for any data, any type, any size, for any type user in any given time frame. Regardless of how big your organization is, the amount of data that you are going to be expected to manage over the next decade is going to grow exponentially.
Today we talk about big data network analytics. A lot of times you think of big data you think of large companies mining almost unmanageable amounts of data. Much of it unstructured data to try to uncover different trends and insights about their customers and their prospects and their products in the market. More data is generated today by machines than by people. The machines on your network, in your infrastructure, routers, switches, servers. Whether you're talking about application of wireless devices, all these things can produce a lot of data. If managed properly that data is the key to get an insight about the health of your IT infrastructure. How things are trending whether it is for capacity planning purposes or just instant health of your network and application performance levels.
Again when I say big data network analytics, this is an issue that really impacts us all. The coming IPv6 represents something big and the need to be able to handle these analytics on your network. I think IPv6 also represents one other thing and it is more personally to you. That’s a new skill challenge and opportunity. This is a chance to get out ahead of something becoming an expert on it. Get experience deploying it in your organization. Deploying IPv6 and get that on your resume and you instantly become extremely valuable to other organizations. It is defiantly a chance to get ahead of things.
Do something that other people aren’t doing that, something that can make you more competitive, makes your organization more competitive. In fact just to illustrate this point we asked in a survey recently hundreds of IT managers. How concerned are you about your team’s ability to proactively find problems and resolve them before end users are impacted? Well, if you look at the results and there's some small numbers and prints and bonds on this chart but, essentially what it saying is 9 out of 10 IT managers out there agree. I’m very concerned about my teams’ ability to proactively find problems and resolve them.
Now that’s our job as IT right, we don’t want to hear from end users complaining, we want to find out about stuff proactively, resolve it before people are impacted. Most people agree on that. Then we moved to another question, and we said in general how well educated and prepared do you feel your IT staff is to handle up and coming IT technology challenges such as IPv6? Well, the results then shifted a little bit. If I could explain what we are looking at there at the right hand side people who answered either barely well-educated or not educated at all, and that was 15% of the people.
People who said I"m somewhat educated on how to handle up and coming IT technology challenges such as IPv6 was 41%. Now let’s be honest with ourselves, when people on a survey answer they're somewhat educated chances are they are not educated enough. You are looking at more than half the respondents. Don’t feel they are prepared from an education stand point to handle upcoming IT technology challengers like IPv6. At the same time 9 out of 10 IT managers say this is something that is important to us, important to get ahead of.
You have that gap there and if you personally can fill that gap and become somebody who gets proactive about it, get educated you can add a lot of value to your organization. Just to equate this to the financial impact it can have on your organization. How much money did US companies loose last year due to network issues? Is it 10 million, 100 million, 500 million, a billion, two billion? What do you think it was, what do you think the answer is? How much did US companies loose last year due to network issues? If you went with a high number two billion you are correct, we try to make this easy one on you.
You always take the highest number in this multiple choices, right? When you go to the baseball game and ask it, "What's today's paid attendance?" You always give the highest number. Same here, $2 billion. Let's put that in practical terms. For your organizations it's probably about 3.6% of your annual revenue on average is lost to network issues, downtime and averages. 3.6% of your revenue.
We're talking about the perfect storm here when it comes to IPv6. You have a major change coming, lack of education, lack of people being proactive about getting out in front of it, and significant financial risk if the deployment isn’t properly monitored and managed to help avoid downtime. That's a little background on the current state of things. What do we do to solve these IPv6 challenges? Well I think you can break it up into four different types of challenges.
You have a challenge with visibility over your entire network and infrastructure, something that's required during deployment. You have a challenge of proper IP address tracking in both the B4 and B6 protocol. Scalability challenges. Not only the scalability of your network but the scalability of the tools used to manage your network and monitor what's going on in your environment.
Finally there's some integration challenges specific to the migration. I mentioned earlier, my guest presenter, co-presenter today Steve Sacchi from A10 network. H's going to talk a little bit about those integration challenges. Real quick, I'm just going to hit on the other three, visibility, IP address tracking and scalability. Let's start with visibility. What needs monitoring in your environment? What are we talking about? You have physical servers; you may have virtual servers, applications, databases, cloud apps. Might be doing voice over IP.
You have the bandwidth and network traffic issues, device health and you have your end user experience and response times. All this needs to be monitored in any IT environment. I mentioned SevOne earlier and what we do essentially that’s our purpose. We have an appliance-based solution we can deploy physically or virtually that you drop on your network and it monitors all those different elements. Out of the box contains all the typical monitoring technologies, things like SNMP and net flow and S flow and IBSLA and all that. We can also bring in any third party data as long as its time stamped third party data, we can bring it in from a reporting perspective.
The result is you get complete visibility of your entire infrastructure from one single pane of glass. It's an out of the box self-contained appliance, self-contained solution, everything you need to monitor your infrastructure. As you go into something like an IPv6 deployment, providing that layer of visibility of what's going on in your environment. It really starts with base lining your current performance. How do you know if you're performing in the IPv6 world if you don’t have any kind of baseline?
Well when you deploy a sub one appliance on your network, it’s going to baseline everything. CPU memory, packets, flow, everything is going to get base lined to very granular detail. Now you can measure against what you will know is normal for specific times of the day and times of the week. With that, you can automatically set thresholds on the high and low ends. If any performance starts to hit closer or exceed the threshold, you can get a proactive alert saying, “There's a trend going on here, whether its capacity issue, a CPU issue whatever, there's a trend going on here and you're going to run into trouble soon.
Having visibility of whether things are trending from an IT performance standpoint in your organization. SevOne also includes functionality such as custom status maps to provide you again we're talking about visibility. Just very graphical quick check visibility of your entire infrastructure in a map. Whether it's a geographic map of a region or a country, whether it’s a map of your data center, you can automatically plot infrastructure objects on this map and then get a quick easy green, yellow, red indicator light to see where the potential trouble issues are upon quick visual inspection.
Bandwidth. Visibility of the bandwidth is another issue. Well, you might say OK. What does IPv6 have to do with bandwidth? Nothing really other than you do need to be able to see who the top talkers are on your network and who the bandwidth hogs are. You need to be able to recognize that whether it's IPv4 or v6.If your network’s coming to its knees and there's a slowdown, who's the culprit? SevOne provides that kind of visibility as well. We talked about IP address tracking.
Whatever tools you're using to monitor, you better make sure that it can handle both v4 and v6 interchangeably side by side, dual stack environment or v4 addresses on the v6 network and or vice versa. SevOne does this. I’m sure we talk about managing IP addresses and IP address tracking. I guarantee it that most of you out there probably at some point use a spreadsheet for this. I think going forward as IPv6, that’s no longer going to be a feasible option. Think about it, you're deploying larger address pools; you have different sub netting techniques, more complex 128 bit actually their small numbers that are not easily readable by the human eye.
More dynamic environments, dynamic address management, spreadsheets aren’t rally the tool; you really need an IP address management tool. That’s something you need to consider in your deployment. Finally, I want to talk about scalability. Talked about visibility, IP address tracking now scalability. There are a lot of tools out there; you guys may be using some of these legacy tools to manage your environment and monitor your IT environment. Here's the problem, a lot of these tools were built for yesterday's networks not today's networks.
They can't handle the volume, the variety, the velocity of data transversity of networks nowadays. A lot of the problems is just with the architecture. These tools were built where you have separate servers, separate hardware for your poling device, your centralized database, maybe a different machine for your reports front end. As your monitor domain grows and expands, people start dropping additional hardware for pollers so you can do more monitoring, monitor more elements.
It becomes a very complex deployment because you're dealing with so many multiple machines. Then what happens is you end up with more full-time employees that need to manage all that hardware. The bottleneck problem occurs because I get all these poplars feeding data up to one centralized database and that becomes a bottleneck and it starts to slow things down. By the time you get to the report front end, you end up getting very slow reports because everything is filtering through these one or two machines here.
It’s simply not scalable for the size and demand of today's networks and nobody's going to be helped by reports that are days old, hours old even, you need today's news now. SevOne on the other hand has solved the scalability problem; we have something called the SevOne cluster. Each one of your appliances can be dropped on a network and if you need to monitor more you can drop another appliance in. They distribute the load among each other in a peer-to-peer environment.
You can go to any of these machines in a pool reporting and they all talk to each other and do their own equal share of the work. The result is you get the reports in seconds and not hours, days or longer. It is the world's most scalable management platform. Again, if you're not familiar with SevOne, you may sure most likely recognize a number of our customers by their logos, Comcast, Xerox, Verizon. We certainly have a lot of customers in the banking and finance space, Thompson Reuters, Morgan Stanley. In the cable industry Time Warner, Cablevision. We have customers in education, in healthcare, federal, local state governments so we span a large breath.
I think what's important here is when you look at each one of the logos on your screen, each one of our customers here. Every one of them had a system in place for monitoring their IT environment. They got to the place when they realized it just wasn’t scalable anymore, it couldn't handle things like IPv6, they just didn’t have the tools built for today's network challenges and they all moved over to use SevOne to monitor aspects of their environment.
Quick question before I turn things over to Steve here. What is tomorrow's news worth to you? Certainly yesterday's news has no value. We're that instant gratification culture right. What is tomorrow's news worth? If I could reach out right now and ask each one of you on this webinar, if you could go magically and grab the newspaper from tomorrow, and get some bit of information out of it and bring that back today, what would you say? I guarantee I can tell you what's going through your minds right now. I want the lotto numbers so when that power jackpot gets up over 500 million again I can play the lottery, maybe some sports scores that you can wager on which we're all very financially motivated people, right?
I actually gave this presentation in Houston the other day, somebody actually raised their hands said, “I’d want to see the obituaries because if I'm going to kick it tomorrow I want to know and have a really good time tonight.” My point is obviously tomorrow's news is incredibly valuable and in the IT monitoring and reporting arena, that’s what SevOne does. It gives you tomorrow's news today. It gives you information that you can act on today to prevent and avert potential disaster or network downtime that could be incredibly financially impacting on your organization.
If there's anything that you take away from the first part of the presentation here in addition to the background on the whole IPv6 landscape. Make sure this, make sure you have the right tool in place to monitor and manage the deployment and the change management process as you go through this. It’s incredibly important if you're going to; again, avoid financial risk. You want to make sure you have visibility of your entire IT infrastructure.
Proactive monitoring of what's going on with real-time dashboards and information of the exact health of your IT status right now. Quick troubleshooting tools, the ability to send proactive alerts so you're getting alerted if anything's trending in the wrong direction outside of a normal, standard deviation from your baseline. it's got to be scalable. Scalability cannot be a roadblock for you guys, OK.
A couple things I just want you to take away from this presentation, if you do want to learn more about SevOne, we do demos online every other week, we call them Demo with Dave with one of our sales engineers. Incredibly knowledgeable resource in the IT monitoring and reporting space. Does different topics every other week, so go to our website sevone.com and look for the Demo with Dave button and sign up for one of our webinars, you can learn a little bit more about what we do and how we can help you.
Also if you want you can go ahead and give SevOne a try, there's a free download on our website to monitor up to 1000 elements on your network. If you want to go ahead and download that, take it for a spin. We'll be happy to talk with you further if you want to learn about SevOne. At this point, I'm going to turn things over to Steve from A10 Networks, and Steve's going to get into a little bit more technical detail around the migration and integration challenges with IPv6. Steve welcome.
Thank you. If you could just go ahead and give control to me there Scott and I'll take control. Fabulous.
OK Steve, you should be set.
OK. Looks like we're good to go. Once again I would like to first of all thank Scott for allowing me to present this part of the seminar. Once again my name is Steve Sacchi. I'm product marketing manager here at A10 Networks. Today's presentation we're going to discuss IPv6 implementation and planning. Are you ready? Real quick a couple of slides just about A10 Networks for those of you who may or may not know who we are. We're considered a leader in application networking
We were founded back in 2004 by a gentleman by the name of Lee Chen. You may not be familiar with his name but some of the other companies that he's founded you're probably familiar with and that was Centillion Networks and Foundry networks which is actually now Brocade Networks. We are headquartered in San Jose California, actually it’s a little bit of an older slide, we actually have over 500 employees worldwide so we're growing pretty consistently.
Today as part of the presentation we'll also discuss our flagship product which is called the AX Series platform, application delivery controller. Our company has been profitable since 2010 with consistent revenue growth from quarter to quarter. With that out of the way, what are the main drivers for migrating to IPv6? I know Scott touched on some of these and of course the first one is IPv6 exhaustion, meaning running out of addresses. Obviously that may be more important depending on which side of the business you're in, meaning that if you're on the enterprise side, your requirement for running out of IPv4 addresses might not be as important as, or I should say as eminent as a service provider who needs to continually hand out addresses and runs up blocks
Adding IPv6 services to legacy IPv4 infrastructure, so we’ll get into this a little bit of how potentially customers or enterprises that have IPv4 infrastructure but would like to provide IPv6 services to either their customers or maybe companies that they’ve merged with or international companies. Those are the two main ones. There are some additional ones but kind of the two biggest onesright there. Of course what's the benefits of IPv6 and of course the ability to handle tremendous amount of devices as Scott alluded to in his presentation.
Of course I've listed some of the devices that you may be familiar with out there, of course the cell phones and smart phones that are now consistently on and have multiple connections over the devices there. I've heard of the story if those of you who may watch discovery when you can’t sleep at night. I heard the story about smart sand where in the future they've designed this technology where they can sprinkle in the highways and this sand that basically connects itself together and builds a network along the side of the highway with sensors and things of that nature.
Once again just let increases to the amount of devices that are out there. Of course IPv6 also brings some additional efficiency the way routers optimize connections and routes and so on. Generally that’s a value to the networking part of the company. The only reason for this slide is if you look down there at the bottom that’s where internet and networking started and then if you look into where it's grown, the 15 billion mark and growing a number of connections and devices listed on there.
Obviously it's growing, everybody knows that. There's also a couple of notes there about Microsoft acquiring Nortel IPv4 address block for $7.5 million. Of course there's now it's been going on over the last years, company buying other company in blocks of address because they are valuable and the growing rate of prices will continue to go up.
Here's how the adoption change worked along. Starting down at the end devices or client end, IPv6 has now been added to windows vista, smart phones, pretty much all iPads, droids. All newer devices traditionally have IPv6 capabilities built right into their operating system. Service providers of course were the first to take advantage of implementing IPv6; they’ve probably been doing pilots and trials for the last three years we've seen. Of course the reason is they have the biggest need for the amount of addresses. Exhaustion is the biggest thing on their mind that’s why they’ve started early.
Enterprise and content providers will also face this problem for various reasons, one of which would be they just want to have IPv6 presence and the other maybe that they may be acquiring companies or merging with companies that already have a very large IPv6 infrastructure. The need to migrate between the two becomes considerably important.
I get this question a lot, what is the ROI? What do I get if I go to IPv6? that's a difficult question to answer because I'm not sure everybody's actually figured it out. There is some reduction in cost long term. I would say the biggest reason is something like this. If you go out to the IPv6 internet and go ahead to www.a10networks.com you'll get our page. As a manufacturer we certainly want to be visible to both sides of the network. What would happen if this happens, right?
You certainly do not want your brand to not be available on the IPv6 portion of the world. This might have considerable value to certain companies, retailers, manufacturers, things of that nature. That has resonant value, the ability to be available on both IPv4 part of the internet as well as IPv6 part of the internet. You don't want to get any of your customers to get a blank screen so obviously this has intrinsic value to you.
I can move along. Let me educate you on the various what we call IPv6 migration technologies. Once again everybody's scared of just jumping from IPv4 to IPv6 otherwise they would have did it already if it was that easy. There are several technologies out there that are available for you to make this migration less painful. The first one there is called dual stack. What dual stack means is you have some kind of router, some kind of application delivery controller such as ours that allows you to run IPv4 and IPv6 independently across the same platform.
The next technologies are what we call the encapsulation technologies. If you look there it's a really good visual where you have an IPv6 packet that rides inside an IPv4 envelope as part of the data stream. Then of course diverse of that where you have an IPv4 packet running inside an IPv6 structure or packet. Then the last one there which is one of the most popular that that we've seen is of course the translation. Being able to actually convert IPv6 to IPv4 using something called carrier grade net or large scale net which we'll get into in the presentation.
Of course here's the problem. You have an IPv4 network, you have an IPv6 network, there's really no compatibility between the two so you need some gateway device such as our application lever controller or some router, gateway that will allow the two to communicate. Of course there's different requirements as you see listed there at home. I don't think anybody really cares whether they have an IPv4 address, an IPv6, or IPv8, as long as you can get to your content and you get to your videos and download your music, you really don't care.
On the enterprise side however, there's some first specific needs and requirements that are which being able, like I said before, to provide maybe an IPv6 connection to IPv6 services but also allow for migration to the existing legacy of IPv4 part of the network or infrastructure. Unless it's the service provider which has the biggest requirement because he needs to continually add addresses. He has to take from the block that he was handed from the INA. Be able to hand out these addresses whether they're IPv or that they're phone provided or phone provider, or are cable TV provider or what have you.
Much roles and terms that you know. Here is the various terms for IPv4 exhaustion, IPv6 migration solutions. LSN CGN. What that stands for is Large Scale NAT, Carrier Grade NAT, and Network Address Translations 444. Large Scale NAT and Carrier Grade NAT mean the same thing nowadays. Traditionally you referred to Large Scale NAT when you were just netting from IP version 4 to IP version 4 and Carrier Grade NAT when you're going from IPv4 to IPv6. I've seen those terms used overlapping. NAT 444 just refers to doing net three times. If you can envision maybe at home you have your Wi-Fi router and you net your private IP address to maybe a public IP address that your provider gave you. Then within his network, he nets you again. that's term 4 to 4 to 4.
SLB-PT stands for Server Load Balancing-Protocol Translation. This is a great example for enterprises where they would convert IPv6 to the right IPv4 infrastructure without changing anything else. NAT 64 just means Network Address Translation from 6 to 4. DNS of course stands DNS look up for 64. DS-Lite stands for Dual-Stack Lite and 6rd which stands for 6 rapid deployment. These both those encapsulation technologies that I refer to. I've got some more diagrams to give you a better example of that. This just shows where some of those features or technology I've shown you might be used. On the service provider side, things like 6rd, Dual Stack Lite on the enterprise side, the SLB-PT and the pure IPv6 to IPv6 load balancing on the enterprise side there.
Here is a couple of diagrams that I have which I really think are useful to help you understand the differences between the technologies. Dual-Stack Lite, the first one. This is basically used by the cable TV provider and you may have lynxes to connect your router at home. Which will basically have this feature built into it. What it will allow in this example if you noticed on the blue, the blue is the IPv4 connection and the orange there is the IPv6 connection. You'll notice that you may have a laptop or a game controller that connects using IPv4 but the service provider has got an IPv6 only connection into your home. The IPv4 is now encapsulated into the IPv6. Goes up the line you'll notice there we have a box called the AX Series. that's basically the Carrier Grade NAT Gateway which will then strip off the IPv6 portion of header and just forward on the IPv4 to the IPv4 internet.
If you look down there on the bottom part of the diagram, it steps you through. 6rd is another encapsulation technology. Basically the same way but just used in reverse. This way the service provider only has the ability given an IPv4 connection into your home but you want to use an IPv6 address. Your IPv6 will then be encapsulated into the IPv4. It will then be stripped off and sent to the IPv6 internet.
Like I said before, a Large Scale NAT or Carrier Grade NAT, what this is now designed to do is used by potentially a service provider that is going to net the IPv4 addresses to the additional public IPv4 addresses. The real goal here is to maximize IPv4 address capacity by doing consolidation. Meaning that if you have 100 addresses, that instead of using a 100 public addresses, you can use say 10 public addresses but they will provide for 100 users across those ten addresses by just netting them. Like I said before, Large Scale NAT also called Carrier Grade NAT is designed to do this.
First I want to give you just an example of what some people refer to as traditional NAT. Carrier Grade NAT is not to be confused with traditional NAT like you do in your home. NAT or network address translation you may do in local router or it just takes couple of addresses that are private conversant to using NAT from across to a public address. Here is the real limitations of traditional NAT.
it's not scalable. Meaning it's designed to only handle say tens of addresses or maybe even 100 of addresses but certainly not 100 of thousands or millions. No application level gateway support. For instance, I don't know if you've ever tried doing running multiple game controllers or Skype or some of those other applications when you're doing too many nets, you'll notice that it won't work all that good. The ability to be complete transparent regardless of what application you use. User quotas, there aren't any meaning there is no way for you to restrict how many ports or how much traffic an individual uses. Of course no logging. These are all limitations of traditional NAT.
Carrier Grade NAT has the capability to do all of these. One of the biggest misconceptions probably spanned by our competition is that NAT that supports a large number of connections is Carrier Grade NAT. that's not true. Just because you could support say 100, 000 connections, doesn't necessarily and you don't do any of the ALG user quotas logging. It doesn't mean you have Carrier Grade NAT.
Here is the first example of where you're using Carrier Grade NAT for IPv6 to IPv4. In the case of where new IPv6 mobile carrier is using IPv6 addressing to all of his mobile users. Still those mobile users still need to get to the IPv4 internet; you'll notice that the connection comes in. IPv6 in the orange once again hits the AX Series and then it's translated during Carrier Grade NAT to IPv4 and then sends that traffic across. You'll also notice the requirements of DNS 64 which run hand in hand.
Meaning that on IP version 6, the DNS request is a quad A request. If you send out that quad A request across an IPv4 internet, you're not going to get any responses. What we do is one of the features as part of NAT64 is we have DNS 64 which won't receive a quad DNS request. We're going to send out both a IPv6 DNS request as both as an IPv4 for real request if you look down on the bottom portion of the diagram there. Whichever response comes back, then obviously that's a response from an IPv6 part of the internet and then it will be serviced a little better. Example of that, of course it's a little bit small.
Server Load Balancing Protocol Translation. When we started out, our product strictly did Server Load Balancing. Server Load Balancing by definition does NAT so it's built into it. If you look at this example here, this might be the example of where I'm a retailer or a manufacturer. I want to make my brand available on the IPv6 portion of the network to offer those services to IPv6 connections. You'll notice that all my infrastructure in blue there in the back and on the right is all IPv4. By providing an IPv6 virtual IP address and then internally will do all the netting from IPv6 to IPv4. This is a quick and easy way for you to provide IPv6 connectivity and services to clients, but not change your infrastructure. Or maybe potentially merge with another company that has IPv6 infrastructure. Once again, this is really designed for a quick migration to IPv6 and allow it to have a quick IPv6 offering.
A couple of examples that I have will really lend itself to the explanation. Here is an example of a mobile provider. A mobile provider had to continue to offer IPv4 addresses to their new mobile users. The problem was that they are running out of addresses. Traditionally they would give a public IPv4 address to each user. In this example here, what they now decided to do was they would assign, if you look in the pictures there, thousands of IPv4 addresses to their private space. They will consolidate those addresses using one or as a fraction of public IPv4 addresses. They are able to get a 10:1 15:1 ratio. That meant that where they may have had a million addresses available, they would only have to be able to support a million customers. Now with that same one million addresses, they can now support 10 to 15 million customers. A significant advantage.
One of the problem was that they couldn't just went out and bought more IPv4 addresses. Going at say $20 a piece, that would be say $20 million for each million that you buy. Certainly that was out of the question for the carrier. By just placing the AX in there, which is in terms of hundreds of thousands of dollars as opposed to millions of dollars, certainly allows to now air those new clients and not incur the exorbitant costs
Here is an example of IP-TV broadcasting video on demand in Japan by NTT Plala. They use the traditional IPv6 to IPv6 or the pure IPv6 all the way across. they're doing an IPv6 to IPv6. A medium server load balancing. they're taking advantageof our IPv6 implementation there so that they can scale very, very large and offer a lot of connections with high availability to all of their customers for a fiber to home network. Kind of like are those that are here in the states have Verizon Fires.
In the last type of deployment there are clients like I said, that require a dual stack scenario where they need to still run an IPv4 and an IPv6 network independently. They don't want to do it using individual products for each of those. They actually used our AX platform which can run both IPv4 and IPv6 on the same platform without any additional cost. That made it very simple for them to add both of networks and then allow them to migrate at their own pace.
Here is the quick example of what I spoke of before we actually were involved in "IPv6 World Day". If you notice, they're on the right. We offered IPv6 services to our network. In the back end of our network we're IPv4 but certainly using SLB-PT, we were able to provide the IPv6 connections so that our website content was visible on the IPv6 part of the network. Once again, the traffic was very small. In retrospect everything else goes on but it was run flawlessly, very easy to implement and actually feel, note down on the bottom we received a zero support calls when we were doing this. Just to show you it's not through vapor ware. There is actually an implementation for these small products still in the rack and we received best of show over to Tokyo and Iraq.
What do you next now that you've heard all this? Probably I'm not telling you anything here that you don't know. Of course the key one there is test the applications. Make sure whatever you're doing is going to work across any kind of NAT and network will have both IPv6 and IPv4. Of course there is a learning curve involved. Of course some training. Certainly you could take advantage of our various IPv4 to IPv6 and NAT migration technologies. You might want to start small like it says there. Just enable a website or a portal using SLB-PT and then maybe run some pilots for IPv6 addressing. If you're an enterprise, you can get familiar with the technology.
A little bit about our product. One thing that makes us unique and I want to show you this slide here. We've actually developed our own operating system. The advantage to this is that other operating systems out there like Linux and UNIX are really not designed to take advantage of Multi-core CPU technologies. Where ours was developed from day one to support Multi-core CPU technologies which allows us to be very efficient. We use a 64-bit scalability. The efficiency of our design allows us to provide much higher performance. that's power and less resource consumption. it's extremely flexible design. Things are easily added to it and scaled up very easily. Here is a couple of our models here.
Of course we have quite a few mature implementations of both Carrier Grade NAT, Large Scale NAT and of course SLB-PT. One of the features that we provide is stateless fail over and hit less upgrades at a very high session rate. Meaning that we can actually provide higher availability implementations. It either comes in a one 1U or 2U platform. The price is extremely low per subscriber. All inclusive basis which means all our features are inclusive. You don't pay site licenses.
This is actually more of a diagram for just traditional server load balancing or application delivery. Where connections come in, they hit either one of our devices here which provides high availability and distributes to the servers in the back end. This provides of course optimization, offloading, also acceleration of course with server load balancing. The real key is to provide a high level of scalability and reliability without the user seeing any kind of delegation of performance. It has a consistent flow all the way through.
Of course on the right we show there is extra recovery of back up data centers capability which would mean if we lost primary center data the backup center date will automatically take over and service that traffic. Here is who is who of some of our customers that we have. I've only noticed that they covers all various industries from large enterprises like Microsoft to service providers carriers, banking institutions, universities, entertainment. We have WWE there and Lucas Films couple of our personal favorites.
Lastly, just a recap. Our ACOS platform is providing the applications delivery controlling server load balancing as well as IPv4 to IPv6 migration. We also have some recommended links you might want to go there. We have some classes on IPv6 readiness or white paper called The End of IPv4 Migration Paths, and several case studies. You can also go to our website and download a version of our software. Once again, it does not have the application acceleration features and capabilities that you'd expect but it certainly has all the features that you've programmed. If you want to run it runs right now in CVM or KVM. that's basically the end of my pitch. I guess at this point I'll pass it back to Scott.
Thanks Steve. Very informative there. I know we're hitting up on the 60-minute mark and we want to respect your time but we also want to give everybody a chance to ask some questions. If you have a question about IPv6, about what SevOne brings to the table, A10Networks and their solutions for migration, please go ahead. Use the Q and A panel. You should see that on the right hand side of your screen. You have to click on that Q and A tab but you can go ahead and text in, chat in a couple of questions. We'll address as many as many of them as we can here.
Steve, I did get a question that came in just a little earlier. Somebody was asking, with your Carrier Grade NAT solution, does that support high viability.
Actually it's pretty unique. What we mean by that of course is that when you go through a routed path and we actually NAT at that session, it's synchronized to either one or multiple AX partners. If for any reason AX connection should be broken or the AX device should fail, then that session will automatically be picked up by another AX and maintain state. The answer is yes and it's actually pretty unique because a lot of our competitors do not provide that capability.
OK. Thank you. Again, just an invite. If anybody has any question, feel free to chat them in via the Q and A panel on the right hand part of your screen. Send us a quick text and we'll address your question. OK, we did just have another one pop up. This one is directed towards SevOne and the topic of scalability. It says, when you mention scalability, how many elements can you monitor from a single appliance. A good question. We can actually monitor within one single SevOne appliance 200,000 elements CPU memory and so on. Incredible scale. that's about just to give you some perspective about four to five times our next competitor. We're an incredible scalable solution. We have customers that monitor a million developments with no degradation to performance. Thank you for that question. Let's just say we'll give it one more second to see if there anybody who wants to chat in a quick Q and A.
All right. Well, with that, I want to thank everybody for their time today. On behalf of myself and Steve with A10Networks, we appreciate you joining us. I hope you were able to learn something. You got some good value added in today's presentation. Of course we invite you to jump out on our website learn more about it and we'll be happy to take any phone calls from you going forward if you have any follow up questions. Enjoy your day everybody.