WAN Virtualization

Virtualization is Driving IT Savings

Virtualization – server virtualization, and to a lesser extent desktop virtualization – is driving IT savings.  This, of course, is almost self-evident to anyone paying attention to what’s going on at the leading edge of computing these days.  Virtualization technology from companies like VMware has taken us from the days of single CPU, single OS to a single CPU with multiple virtual machines towards cloud computing / utility computing, leveraging the efficient pooling of on-demand server infrastructure to scale to run an almost arbitrary set of applications – be it public cloud computing, private cloud computing, or hybrid cloud computing. 

After talks with a number of industry analysts, we recently settled on WAN Virtualization as the term to describe the new market category that Talari is pioneering with our Adaptive Private Networking technology.

I think the term works well on a couple of levels.  For one, it's in the same vein as, yet distinct from, WAN Optimization, befitting the complementary nature of WAN Virtualization and WAN Optimization technology.  More importantly, just as Enterprise WANs have gone from being composed of point-to-point leased lines, to leveraging reliable public cloud networking (X.25, Frame Relay, ATM and now MPLS) and a bit of unreliable public cloud networking (the public Internet) for remote access and IPSec VPN backup connectivity, WAN Virtualization technology like Talari's is changing enterprise WANs in the same way that VMware and server virtualization are changing enterprise computing, enabling reliable public cloud networking.

WAN Virtualization leverages the economics of the public cloud network that is the Internet, as well as enabling hybrid cloud networking utilizing a combination of private WAN connectivity and public Internet connectivity.

If you've been reading this blog for a while, you've seen previous posts here and here on how Talari's technology is similar to RAID. Since RAID was, after all, essentially storage virtualization phase 1 – or phase 0, I suppose, depending on your point of view – the term works at that level as well.

In future posts, you'll be hearing a lot more from us about the exciting possibilities that result from synergies across WAN Virtualization, server virtualization and WAN Optimization.

APN's Innovations to Improve WAN Application Performance

In my first post in this trilogy, we looked at the factors which affect application performance on the WAN. In the second, we enumerated the various techniques vendors have implemented historically to address these issues.  Here, we'll take a look at what Talari’s Adaptive Private Networking (APN) solution brings to the table, and how these new techniques complement the existing solutions that the WAN Optimization, Virtual Desktop and WAFS/caching vendors deliver today.  

In addition to doing bandwidth management/QoS for outbound traffic essentially as any modern middlebox does, and offering TCP termination to address the bandwidth-delay product issue as many existing solutions do, APN innovates in a number of areas in addressing WAN application performance.

Most obviously, of course, APN's Resilient Multipath Connectivity enables link aggregation of diverse IP WAN connections while still delivering predictable and reliable performance.  Inexpensive Internet connections can now be used to reliably augment and/or replace expensive WAN connections like MPLS.

APN also offers innovation in a number of other application performance and performance predictability areas as well.  APN’s techniques for reliable packet delivery mitigate the effects of loss on TCP applications, buffering and retransmitting packets to make the WAN look like a zero-loss network, with occasional bouts of jitter.  Of course, the jitter introduced is no worse than a TCP application on the WAN has to be able to handle in case a packet is stuck in a queue in some WAN router behind a congested WAN link.  This enables TCP window sizes to stay at maximum, without negative impact on the WAN itself.  Further, in the face of more than moderate loss, or in the face of bursts of jitter, APN will move traffic, especially timing sensitive traffic, to a better path with lower loss, and/or latency.  By making this move away from “flaky” bandwidth sub-second (< ~250 ms across the continental U.S, < ~600 ms even for traffic going halfway around the world), APN delivers reliable predictable application performance even while using the “pretty good but not business quality" public Internet for its sources of bandwidth.

For real-time traffic like VoIP or videoconferencing applications (or even low-bandwidth, time-sensitive interactive applications like Citrix), APN will first put the real-time traffic on the best path with the lowest loss and lowest jitter to the destination.  As per above, if loss or congestion (jitter) is detected, it will switch the traffic, sub-second,  to a better path.  But for “platinum” quality voice – or for that matter, platinum quality videoconferencing, if you’ve got the bandwidth – it will send a replica of the packet stream on a different paths, as unrelated in terms of local and remote WAN links as possible; the replica is discarded at the receiving APN appliance.  Routing, of course, has always taken care of up/down network availability issues.  With this technique, even if what is usually the best performing path suddenly starts experiencing 68% loss or even a 150 ms burst of jitter, the same packet will show up at the destination along the other path, a few milliseconds later, and the application will never miss a beat.  APN counts on there being jitter buffers in the receiving hosts for this scheme to work, and in fact this is exactly how VoIP and videoconferencing work.

On the bandwidth mgmt/QoS front, APN delivers superior results for outbound traffic over existing solutions because, as a two-ended solution which continuously knows the state of all the network paths to the destination, it gets priority packets not just out the local link first – which is all existing solutions in practice do, since RSVP and IntServ never really “happened” – but to the other end of the WAN first, which is what you really want.

For inbound WAN bandwidth mgmt, in addition to doing a cruder version of what Packeteer PacketShapers have always done for limiting generic public Internet traffic on links shared with APN connections, it does far more sophisticated inbound management than any prior IP forwarding devices for inbound traffic coming from other APN sites, avoiding most last mile-based loss and congestion in the first place.  I.e. unlike TCP, which, as one sees with the famous TCP sawtooth, is actually designed to cause congestion and loss before backing off, APN will sometime temporarily attempt to send 3.1 or 3.2 Mbps into a 3 Mbps link, say, but will never try to offer 5 or 6 Mbps of load to the link, as standard TCP will habitually do, especially if there is inbound traffic from two or more locations.

WAN Optimization products, as a whole, combat the effects of limited bandwidth and high average latency.  WAN Opt typically delivers a 2x – 4x bandwidth benefit by freeing up bandwidth with its compression/caching techniques.  WAN Opt addresses high average latency with its generic TCP termination capabilities, and with its application-specific chattiness reduction optimizations for CIFS and HTTP.  WAFS and caching products generally combat the speed-of-light latency problem as well, in a less general, more focused fashion, but have historically achieved less success in the marketplace, in large part because of the extremely limited bandwidth available at most smaller locations.

Adaptive Private Networking is quite complementary to WAN Optimization, since it not only cost effectively enables much more bandwidth, but excels at combating the effects of packet loss and jitter (i.e. worst case latency), as well as frequently preventing loss and high jitter from occurring at the last-mile in the first place.  APN is also complementary – indeed, even an enabling technology – for distributed, replicated file services by enabling the provision of large amounts of inexpensive bandwidth, a solution not heretofore cost effective.

WAN Opt, WAFS, caching, Citrix and other application-specific solutions help some for bandwidth and are great at addressing speed-of-light latency issues; APN, while doing little to address speed-of-light issues, is unique in its approach to addressing the effects of WAN packet loss and congestion-based latency/jitter, as well as in preventing much of that performance-killing packet loss and congestion from occurring in the first place.

The Whole is Greater than the Sum of the Parts

In Peter Sevcik's and Rebecca Wetzel's recent report on Flex-IT Networking: An Emerging Trend, they illustrate how the the confluence of some key technologies and capabilities will transform the way wide area networks will be built in the future.

Obviously, my interest in this area is directly related to the fact that one of the capabilities is to build a reliable private WAN out of a 'mosaic of network transport options', to create an 'Integrated Transport Network (ITN)'. This is exactly what APN technology was designed to do.

Server consolidation accelerated by server virtualization continues to be a hot trend and provides real OpEx and CapEx savings for IT departments. However, as computing resources become centralized and more distant from the end user, the bandwidth and latency demands on the enterprise WAN have increased significantly.  This paper points out that that colocation facilities combined with an ITN are primed to become an integral part of your WAN design to help solve this problem.

Colocation facilities, that have traditionally been used to house Internet facing web services and for disaster recovery, will play a broader role with an ITN approach.  Now they can be used to inexpensively consolidate branch services such as an email server, firewall and intrusion detection system within a geographical region close to the users without the need for an expensive private WAN connection. Connected with plentiful and cheap bandwidth they also provide an answer to the age old question of whether to back-haul your Internet traffic or not.

The resulting solution will provide the branches with both more Intranet and Internet bandwidth which is great for the users, and for the IT department it reduces costs and the number of locations that need to be managed.

So when you add it all up, cheap bandwidth allows you to move services to where they are more cost effective to manage, virtualization allows you to rent less rack space in those facilities and an ITN powered by APN technology allows all of the inexpensive network resources you can leverage to run even more reliably than the MPLS you thought you needed.

WAN Service History and Oligopoly Margins (or, Why does MPLS cost so much more than Internet connectivity?)

by Talari Blogger-in-Chief Andy Gottlieb...

MPLS is really expensive.  Internet connectivity is really inexpensive.

Why do private WANs cost so much?  More specifically, why are enterprise WAN services (MPLS, Frame Relay) priced so high?

Let’s take a look at the history of why private WAN services are priced the way they are.

I double majored in economics as an undergraduate, and I'm a firm believer that the market sets prices, and those prices are based on supply and demand.

Why are private WANs priced as high as they are?  It’s clearly not because of cost-based pricing.  Let’s assume that in the early 1990s the customer price for Frame Relay service quite likely was largely a function of the cost of providing the service.  Back then, in fact, Frame Relay was ground breaking price/performance, far better than the X.25 and leased lines it replaced.

In the roughly 12-15 years since that service became widespread, the cost/bit of the equipment to operate a carrier data network has gone down dramatically.  And while the ongoing costs of an “over-engineered” IP network, and of the headcount needed to provide better MTTRs in enterprise SLAs for last mile failures, mean that the cost/bit of running a private WAN is somewhat higher than for “plain old Internet service”, those ongoing costs are certainly not more than about, say, twice the cost/bit of providing Internet service.

I noted in an earlier post that MPLS is priced at $450 - $1,500 / Mbps per month for the copper connectivity typically deployed at all but the very largest enterprise locations, while the monthly price of broadband connectivity is ~$3 - $15 / Mbps per month. 

Fiber MPLS connectivity is typically in the $100 - $300 / Mbps per month range, while fiber Internet connectivity at a customer location is typically half that, and high bandwidth Internet connectivity at a colo facility is typically down around $10 - $20 / Mbps per month.

There is simply no way that the cost of providing private WAN connectivity is 30 - 100 times the cost of providing public Internet connectivity.  And so the pricing is clearly based on what the market will bear.  Which is surely the right of the carriers – and should be – in a free market economy.  But a look deeper into why the pricing is what it is can be quite revealing.

 

Broadband Internet service costs roughly $25 - $200 per month, and the bandwidth available is in the range of 3 Mbps to 50 Mbps.  One can argue that the telecom SPs were dragged into pricing this way because the cable companies – which had no telecom service market to protect – priced their product somewhat aggressively to target the consumer market, and have continued to increase the bandwidth available (e.g. 50 Mbps downstream for $189.95 / month from Comcast: http://www.comcast.com/About/PressRelease/PressReleaseDetail.ashx?PRID=814)

Whatever the initial reason for the pricing strategy, the fact remains that broadband is a growth business for the telecom SPs, and outside of wireless, broadband is one of their biggest investment areas.  And while monthly service prices for broadband seem to have more or less stabilized, price/bit continues to improve as the amount of bandwidth available per connection increases.  The price for Frame Relay and MPLS services (MPLS has always been priced under the Frame Relay umbrella, and is almost always within ~25% of the price of FR service of equivalent bandwidth) has come down somewhat in the last few years after being fairly flat for the first half of this decade, but has still not come down with Moore’s Law as with, oh... everything else in technology! 

Why is this?  I’d submit that it’s primarily for the following reasons:

• The last mile RBOC/PTT monopoly  “Backhoes don’t obey Moore’s Law” is the old saying, and with the Telcos owning the last mile TDM links – especially the copper connections to the overwhelming majority of business locations – they have been able to continue charging monopoly rents for these links.  Other than at colo facilities and in portions of some “NFL cities”, the same has largely been true for higher speed fiber links to business locations as well.

• Oligopoly of small number of vendors as credible national (global) providers of FR, MPLS  The oligopoly of 2 ½ vendors (AT&T, Verizon [formerly MCI], plus Sprint as the ½) in the U.S.,  and similar monopolies or duopolies in most other countries, limited enterprise’s choice of providers, especially since there has never been meaningful interoperability of FR or MPLS networks.  With fewer choices comes higher pricing.

• WAN buyers are correctly risk averse  Given limited IT staffs at most remote locations, it makes sense to be risk averse when it comes to the WAN, as the yearly savings of going with a cheaper but less reliable WAN service can sometimes be eaten up in handling even a single incident.  And given limited staffs, paying lower WAN costs and handling connectivity problems yourself when they arise is not even an option for some businesses with large numbers of locations.

• …and so they are not hugely price elastic  Unlike consumers, and many other parts of the computing or networking business, where a price reduction of 2x can sometimes result in 10 times the sales, service providers found that lowering private WAN prices did not result in higher revenues.  So they didn’t lower prices!

• The Internet on its own never got reliable enough  While public Internet connections have improved from the ~1 nine (90% - 95%) reliability they had 10 years ago, they are still only about 2 nines (99%) reliable – where reliability means the union of simple connection availability plus the ability to get your packets through to their destination without being lost or excessively delayed – unlike the 3½ to 4 nines reliability available from FR and MPLS.

• …and unlikely ever will The reasons for this have little to do with technology, and everything to do with economics and billing.  The carriers simply don’t get paid extra for making the public Internet – a network of networks – 4 nines reliable, and so it doesn’t pay for them to over-engineer their networks, and the network interconnections, and supporting network management, troubleshooting, etc. to do so.  The billing issues would be incredibly complex as well.  The billing issue alone is a key reason why technologies like multicast and IntServ over the public Internet never really “happened” either.

• Carriers are milking the FR/MPLS cash cow  While the telecom SPs have indeed aggressively moved customers from FR to MPLS, under the auspices of the lower cost of operating a single data network and ostensibly to deliver the QoS needed for real time applications like VoIP, they haven’t been investing all that much in their enterprise data networks, but instead have focused their investments in wireless, consumer broadband and perhaps the Internet core.

Given the above factors, absent a technology like Adaptive Private Networking (APN), I’d argue that Enterprise WAN managers – even those generally aggressive in rolling out new IT for competitive advantage or to lower costs – have done exactly the right thing being conservative with their WANs for the past 10 years.

But just as RAID enabled businesses to leverage the PC hard disk for their enterprise-class storage needs, APN, by solving the reliability and performance predictability issues associated with using public Internet transport, will enable enterprises to finally take advantage of the carrier's huge and ongoing investments in the public Internet – and indeed, in 4G wireless technology – for their corporate Intranets.

…and finally gain leverage over their telecom SPs and do something about those extremely high prices.

 The market does work!

 

APN and RAID: the Technical Commonalities

by Talari Blogger-in-Chief Andy Gottlieb...

As I note whenever I’m asked what it is that Talari is bringing to Enterprise networks, Adaptive Private Networking is doing for WANs what RAID did for storage. 

In a previous post, we looked at the analogy to RAID at the business benefits level.  Here, we examine the parallels on the technical level.

Wrapping hardware and intelligent software around core enabling technology

Seagate hard disk technology was originally designed for and targeted at the nascent but fast growing market for personal computers. Each hard disk was not nearly as reliable, having nowhere near the MTBF or seek times, as the existing mainframe/minicomputer disks.  But RAID - Redundant Array of Inexpensive Disks - took advantage of the massively better price/bit of the Seagate hard disk to revolutionize the storage market. By combining a layer of hardware and intelligent software with multiple of these inexpensive disks, RAID delivered a system with higher capacity, lower cost, competitive – and ultimately superior – access times, and greater reliability than the older generation storage solutions.

For APN, the analogous enabling technology is the public Internet.  By wrapping a two-ended system of appliance-based hardware running intelligent software around multiple WAN connections – most or all of which are Internet connections – APN creates an enterprise WAN which is lower cost, massively better cost/bit, higher capacity, and – critically – more reliable than the best single vendor MPLS WAN.

Redundancy to deliver application continuity

The basic idea behind RAID – and behind APN as well – is that while two devices operating in series which each have 99% reliability will deliver a system with only .99 *.99 = 98% reliability, a properly designed system with the same two devices operating in parallel  will deliver 1 – (1- .99) * (1 - .99) = 99.99% reliability.

The key phrase, of course, is proper design.

The first premise of RAID is that the loss of any single disk ensures not only that no data is lost, but that the application – data reads and writes – continues to function normally, without meaningful performance degradation.  This is essentially what RAID Level 1 delivered.

Existing traditional routed networks with appropriate link and device redundancy at each location, of course, provide network availability in the face of any hard single link failure or router failure.  But this "no loss of connectivity" is merely the equivalent of "no data loss" in the storage world.  Even a no-single-point-of-failure routed network does not provide for application continuity in all cases, as a routed network can take upwards of 30 seconds at times for router convergence in the face of a given failure.  More importantly, routing does not handle the case where packet loss or excessive latency causes significant problems with application performance.  Yet these “soft failures”, due to congestion on shared IP networks, especially Wide Area Networks, occur with far greater frequency than hard link or device failures.  

In fact, it is precisely because of the congestion-based packet loss and jitter which occurs at Internet peering points between ISPs that the public Internet has earned its "works pretty well most of the time" reputation.  Of course, "pretty well most of the time" just ain't good enough for your Enterprise WAN.

Handling "soft failures" - and doing it quickly!

APN, by leveraging multiple paths across the network between locations, ensures that no single hard or soft failure of a network link, device or peering point in the middle of the Internet will cause a loss of connectivity or application predictability.

APN does its equivalent of RAID Level 1 by continuous measurement of network path performance (loss, jitter, latency, bandwidth) and sub-second response to problems with any network path. In typically less than 3 round-trip times (RTTs), APN will move traffic off of a path experiencing high loss or excessive jitter.

For TCP applications, APN also delivers this predictability by buffering packets from flows and retransmitting them in the face of loss. 

For real-time application flows, APN can replicate the traffic on multiple paths, suppressing the duplicate packets at the receiving appliance, providing almost an exact equivalent mechanism to what RAID 1 does.  Application predictability, by avoiding loss and minimizing jitter, is much more important than efficient capacity utilization for real-time apps like VoIP.  By using Internet connections, the cost of that extra bandwidth is minimal.  For VoIP, it’s actually trivial.  Even for videoconferencing, it’s fairly small where the capacity exists.  [Note that APN also allows “conditional” replication of such flows, based on availability of bandwidth at the time.]

Striping = greater throughput

RAID Levels 2 – 5 (and up) allow improved data access performance by doing bit, byte and/or block level striping across multiple disks, allowing simultaneous disk seeks.  With APN, an individual TCP flow can be striped across multiple paths/links to deliver greater aggregate throughput.  The buffering and retransmission of packets, in conjunction with buffering and packet reordering at the receiving APN appliance and packet forwarding logic armed with the relative latencies of each of the network paths involved, enables high throughput even in the face of packet loss.

Automating the solution to MTBF and MTTR issues 

Network availability and predictability are quite naturally the number one concern of any enterprise WAN manager.  Just as RAID changed how storage systems were designed, an APN-based WAN turns the historic importance of, and so emphasis on, certain metrics on their head.

Before RAID, the MTBF of a disk subsystem was a critical factor in designing a highly reliable and available IT system.  With RAID, this concern pretty much vanished. Before RAID, MTTR of the storage system was a big deal as well. RAID solutions allowed much faster system MTTR (just swap out the hard disk and insert a relatively cheap replacement disk). The defective disk itself is almost never actually “repaired” any longer.  And with proper redundant system-wide design and automatic synchronization and backup processes in place, if in fact the MTTR of even making that disk swap ends up being 24 or 48 hours, no one is particularly bothered.

Similarly, prior to APN, the union of the network availability and the predictability of packet delivery of the private WAN, usually somewhere in the 99.95% - 99.99% range, has quite correctly been of huge importance to the enterprise network manager.  And the private WAN provider's SLA for a 4 hour MTTR, say, when a WAN problem does occur has also been extremely important.

With APN, however, the need for each WAN “subsystem” to be 99.9%+ predictable and reliable is greatly reduced. Given multiple diverse WAN connections combined with sub-second switchover in the case of packet delivery problems with any of them, the need for a 4 hour MTTR guarantee goes away as well.  Who cares if the MTTR is 24 or even 48 hours for a broadband connection at a branch office, say, if the network continues to work, and users notice no loss in connectivity and minimal loss in application performance and predictability?  And if delivering application predictability even during those rare times when a given link is down completely is important for your application or location, by having 3 diverse connections at that location, rather than 2, you will have superior application performance as well as network and application availability versus a private WAN actually delivering 99.99% uptime with an SLA committing to a 1 hour MTTR!

 

RAID revolutionized storage economics, improving storage reliability and capacity while radically reducing costs.  In a very similar way, APN is revolutionizing enterprise WAN economics, improving WAN reliability and enabling substantial capacity increases while simultaneously enabling network managers for the first time in more than a decade to not just have control over their WAN costs and their WAN providers, but to radically reduce that huge budget item as well.

Report: The Compelling ROI of Adaptive Private Networking

In his new white paper, The Compelling ROI of Adaptive Private Networking, well-known industry analyst and Network World columnist Dr. Jim Metzler of Ashton, Metzler & Associates analyzes data from two companies and finds that APN delivers:

• 45% – 86% monthly WAN expense reduction
• Payback in 3 – 7 months
• Bandwidth to each site increased between two and twelve times versus their current network
• End-to-End QoS functionality
• Greater availability and reliability than typical Frame Relay or MPLS services

Check out The Compelling ROI of Adaptive Private Networking.

The RAID Analogy

by Talari Blogger-in-Chief Andy Gottlieb...

When I’m asked for the “elevator pitch” on what Adaptive Private Networking is, I always go back to the RAID analogy – “APN is doing for WANs what RAID did for storage”.  Since we’re not on an elevator, let’s examine this analogy in some detail.

The analogy works both at the business benefits level, and at the technical level.  We’ll look primarily at the benefits part in this post, and the technical parallels in a future post.

Before RAID, in the 1970s and the 1980s there was a large market for expensive mainframe attached storage systems, which IBM dominated, and similar markets for minicomputer storage.  IBM’s margins were extremely high, and even the vendors of “plug compatible” storage systems had high gross margins as well.  They competed on the benefits of their “proprietary disk drive technology”, and only within a narrow range (a factor of 2 or 2 ½ at most) on cost.

Along came Seagate hard disk technology, targeted at the nascent but fast growing market for personal computers. Each hard disk was not nearly as reliable, having nowhere near the MTBF or seek times, of the mainframe/mini disks, but were built in high volumes with massively better price/bit.

RAID - Redundant Array of Inexpensive Disks took advantage of that Seagate disk price/bit to revolutionize the storage market. RAID solved the reliability and capacity limitations of the PC hard disk, and enabled them to be leveraged for data center business quality storage systems.  With RAID technology, companies like EMC built bigger, faster, cheaper and more reliable storage subsystems.  Within a very short time, no one tried to build big, fast reliable proprietary single disk storage systems any longer.

For APN, the existing high margin market is that for private WANs, which today means MPLS and Frame Relay.  Where IBM was the single dominant storage vendor before RAID, today with private WANs, the oligopoly of AT&T and Verizon (and to a much smaller extent, Sprint) in the United States, and government-sponsored telecom monopolies or near monopolies in other countries dominate the market, with 1 or 2 vendors in each company controlling 70% or more market share for these services.

Pricing of private WAN service, which was somewhat aggressive and clearly related to the cost of the service when Frame Relay was introduced in the early 1990s, now bears no correlation whatsoever to the cost of providing the service.  Rather, due to the last mile access (copper and fiber) monopoly by the RBOCs and the oligopoly of nationwide / global private WAN service providers, the price / bit of these services has barely come down in the last 10 years, even as the price of almost everything else has followed the classic Moore’s Law curve downward.

The irony here is that the analogous enabling technology for APN is the public Internet – much of it provided by these very same service providers!  Just as with the PC and the PC hard disk in the 1980s, massive investment in the Internet and broadband has been targeted at individuals and small business, and given the combination of high volumes and competition in that market, the price/bit here has continued to come down with Moore’s Law.  Internet access now costs $3 - $15 / Mbps per month, versus the $450 - $1500 / Mbps per month which the telecom SPs charge for MPLS.

While Talari’s APN technology is pretty sophisticated in terms of how we do real-time per packet measurements and sub-second response – more on that next time – just as with RAID, the price/performance comes from providing the surrounding reliability to enable leverage of Internet economics.  Where RAID wrapped a layer of hardware and intelligent software around multiple PC hard disks, APN’s two-ended, appliance-based solution does something similar with multiple WAN connections - existing private WANs and high speed Internet connections (T3, OC3, Metro Ethernet, etc.) at data centers and large sites, as well as existing connections and any type of broadband Internet links for branch/smaller locations.

RAID leveraged the PC hard disk technology to revolutionize business storage cost, capacity and reliability.  APN leverages the most powerful, ubiquitous, low cost communications network ever created – the public Internet – to deliver Enterprise WANs that are far higher bandwidth, far lower cost and more reliable than the best proprietary single service provider WANs available today.

It may well take a bit longer for APN to change the networking world as completely as RAID changed the storage world, but the parallels in terms of surrounding superior economics with business-class reliability are striking.

What would you do with 20 times the branch bandwidth?

by Talari Blogger-in-Chief Andy Gottlieb...

Fish don’t notice water.  It’s just a given of their environment.

Similarly, scarcity of enterprise WAN bandwidth is a "fact" that IT managers have long taken as a given.  Despite 10 Gbps LAN backbone and server links today, and a minimum of a switched 100 Mbps Fast Ethernet link to every desktop, with Gigabit Ethernet desktop connections more and more frequent, network managers – and application developers, too! – live with the reality that a 1.5 Mbps T1 WAN connection shared across 5 to 50 people is frequently all that’s available.  Even at larger sites, average WAN bandwidth per person will often be similar (or even lower). 

This 3 or 4 order-of-magnitude difference in the availability of bandwidth on the WAN has major implications for network design, server deployment and applications which can be run on the corporate WAN.  And in tough economic times like these, spending twice the amount on your WAN to get more bandwidth to run applications is just not an option.  Heck, spending the same amount you did last year is tough for some companies to do.

When we talk to customers about Adaptive Private Networking (APN) and its benefits, some customers pick up on this idea that they can now afford to have 20 or 30 times the WAN bandwidth they’ve been forced to live with, but for many, it’s an idea that’s so foreign they’ve not though much about the possibilities.

A great example of an application APN enables is videoconferencing.  “Gee, I’d never run ‘real’ videoconferences over the public Internet” is what any network manager worth his salt knows, after all.  That’s what QoS on MPLS is for. 

At Interop Las Vegas 2009, we’ll be demonstrating how having a lot of bandwidth changes the way networks are designed and applications are run.  While almost all techniques from WAN Optimization vendors like Riverbed and Blue Coat are about conserving bandwidth, APN actually takes the opposite approach, focusing, much as RAID did with “cheap” PC hard disks for storage, on making inexpensive bits business quality. 

Because of the inexpensive nature of Internet bandwidth, not only does APN make it affordable to have enough bandwidth to run high definition videoconferencing to the smallest of branch offices simultaneously on your now “converged” network with the rest of your data traffic, it actually duplicates videoconferencing and VoIP sessions to enable very high quality even over “pretty good but by no means perfect” public Internet connections. [Note: such duplication is administrator configurable, and can also be done dependent on the availability of bandwidth.]

So if you’re in Las Vegas next week, please stop by booth 1567 and see for yourself the video quality APN enables even in the face of network conditions which would ruin a simple voice call.

And start thinking about the applications you could run, and the centralization redesigns which would be possible, if only you had 20 times the bandwidth on your WAN. 

Because now you can afford to.

Why is this the right time for an Enterprise WAN revolution?

by Talari Blogger-in-Chief Andy Gottlieb...

Welcome.

Talari is doing for Enterprise Wide Area Networks what RAID did for storage – delivering a network with 30 – 100x bandwidth/$, monthly WAN costs reduced by 40% - 90%, and with greater reliability than existing private WANs using Frame Relay or MPLS.

In a carrier pricing environment where a price/bit factor of 2x is enormous, Adaptive Private Networking brings Moore’s Law and Internet economics to Enterprise WAN buyers for the first time in 15 years.

To answer the question in the title, I’d like to start by telling you a story.

It's a story about the market for mainframe computers, with obscenely high margins and dominated by a single vendor, and later minicomputers with a handful of vendors and very high margins as well. Along came the inexpensive personal computer, and not much happened to the proprietary mainframe and minicomputer business – until those “consumer” PCs were connected by LANs, interconnected by WANs, augmented by client-server computing and reliable distributed software – and the PC hardware / software and LAN/WAN industries grew a huge market, built an ecosystem, and took huge dollars out of the hide of the mainframe and minicomputer budgets.

Wait – that story has already played out…

Ok, let’s talk about the market for expensive mainframe attached storage systems, “plug compatible” with their own extremely high margins and “proprietary disk drive technology” and dominated by a single vendor (IBM). Along came Seagate PC hard disk technology. Not much happened at first. But then along came RAID - Redundant Array of Inexpensive Disks. With RAID technology, companies like EMC built bigger, faster, cheaper and more reliable storage subsystems, and built multi-billion dollar businesses leveraging this “consumer” hard disk technology, and grew a huge market and ecosystem out of the hide of the proprietary mainframe and minicomputer storage budgets. Today, no one tries to build big, fast reliable proprietary single disk storage systems any longer.

Hold on – that story has played out, too…

Actually, this is a story about corporate WANs and the extremely high margins of an industry - the telecom SPs - dominated by one or two vendors in each country. And then along came the Internet and broadband, and… well, in fact all these years later, nothing much has happened to the base corporate WAN business. Sure, Internet connectivity and broadband (cable, DSL) have become a large market, first for surfing the web and email, then e-commerce, etc. Internet broadband is actually the growth business for the telecom SPs, and they continue to invest heavily in these networks, focusing on consumer adoption rather than on enterprise needs.

VPN technology came along and business began to use the Internet for remote access, and in some cases for backup connectivity. But private Frame Relay and MPLS WANs, with per megabit prices of $450 - $1500 per month, remain the dominant way to build corporate Intranets, and are just about the last holdouts from the pressures of Moore’s Law and Internet economics. They are a $25B annual business, including $15B+ for higher bandwidth (fract-T1 and above) connectivity.

VPNs make the Internet secure, and Internet and broadband performance and reliability have gotten a lot better over the last several years. But in fact Internet connectivity on its own does not meet the 3½ nines (99.95%), let alone four nines (99.99%), standard of reliability that businesses have come to expect of their Enterprise WANs. And so even though there is now literally a 100x difference in the price/bit of broadband – at $3 - $15 / Mbps – versus the price of private WAN links, very few businesses, especially larger businesses, have moved off of those expensive, proprietary networks they’ve been using for years, despite the fact that pricing has stayed so high, and bandwidth at most branch offices is less than what’s available to the average home broadband user.

Enter Talari Networks and Adaptive Private Networking (APN).

Just as RAID wrapped a layer of hardware and intelligent software around Seagate hard disk technology, APN does something similar with multiple WAN connections - existing private WANs as well as any kind of Internet WAN links, be they DSL, cable, fiber, Metro Ethernet, WiFi/WiMax, etc.

APN technology uses end-end algorithms to do dynamic, real-time, per-packet traffic engineering.

Rather than optimize each application for the network, APN optimizes the network fabric for all applications.

 APN enables businesses to build a WAN that is simultaneously less expensive, much higher bandwidth, much lower cost and with lower ongoing operational costs than today’s proprietary, single vendor WANs.
APN solves the reliability problem with running VoIP safely on your IP WAN, and the cost and reliability issues with running videoconferencing on your IP WAN.

 And APN can be added seamlessly into existing networks, rather than requiring a forklift upgrade per site, as migration between services even from one’s existing carrier can require.

So APN unlocks the stranglehold that AT&T and Verizon have in this country on their customers’ WANs and WAN budgets. And it does this by leveraging those same carriers (and the cable folks, too) “consumer” broadband networks.

The current WAN Optimization industry, led by Riverbed, is built on the long-standing assumption that your existing network is as reliable as you need it to be, but bits are expensive – and the need, therefore, is to squeeze as much out of those thin pipes as you can.

The WAN Optimization companies are focused on using hard disks to do compression, and the issues of getting bits onto and off of hard disks scalably and at high enough rates. They’re also focused on doing application-specific optimizations to deal with the fact that there is so little available bandwidth on today’s corporate WANs, and on selling the soft benefits of “application acceleration”, since they quickly learned that the 2x to maybe 4x price/bit benefit they deliver over existing networks isn’t a compelling enough reason alone for most customers to deploy it everywhere.

Talari’s insight is that in fact, there are lots of cheap bits out there - they just need to be made reliable enough to be business quality.

A 2x change rarely moves markets, but an order-of-magnitude-plus change does.

With Adaptive Private Networking, Talari intends to build a very big business – for ourselves and our ecosystem partners – out of the hide of that $15B+ Enterprise WAN service budget, by enabling customers to attain huge savings on one of the very biggest line items in their monthly IT budget, 2 orders of magnitude bandwidth per dollar, and equal and in fact better reliability and application predictability than their current locked-in WAN delivers.

More – a lot more – to come.