We’re trying a little something new – and well, if you’re going to try something new, try something new on a Friday.

Network Performance Daily, as you know, is the company blog of NetQoS. We try to be informative, but while our marketing material implies that all we ever do is worry about network performance, we also take the time out to be human beings.

So we came up with this idea: “Profiles in Performance” to try to give you some insight into the people behind the technology. We start with Carol Schiraldi. Wicked programmer, yes, but also gifted fine art photographer. We hope you’ll enjoy it.

In May, 1998, Stephen Glass, who then worked at The New Republic, wrote an article called called “Hack Heaven,” about a 15-year old hacker named Ian Restil.  According to that story, Ian Restil used a computer at his high school library to hack into software firm “Jukt Micronics.”  Jukt decided it would be cheaper to hire Restil to tell them how he did it rather than have their in-house engineers determine how he did.  Glass claimed that stories like Restil’s were “common” and that “Computer Insider,” a newsletter for hackers, estimated that 900 hackers were hired.

It was a compelling story, and one which resonated with the 1998 audience of The New Republic – the idea of hacker protection rackets.  Except, none of it was true.  Restil was fiction.  Jukt Micronics was fiction.  Computer Insider was fiction.  There was no “Center for Interstate Online Investigations,” no radio advertisement against hacker protection in Nevada, no “Uniform Computer Security Act,” no “National Assembly of Hackers.” Even Jukt Micronics Web site was a (pathetic) fake one set up by Glass on members.aol.com.    This was revealed by Adam Penenberg, then working at Forbes Digital, (a milestone for internet journalism – as an online news site took down the star reporter of one of the most storied print magazine publications.) 

I mention this story, because that story bears a bit of a resemblance to this one, published by the Associated Press on Mar. 25th: “Teen Hacker turns corporate cyber-crime consultant.”

WELLINGTON, New Zealand - A New Zealand teenager who helped a crime gang hack into more than 1 million computers worldwide and skim millions of dollars from bank accounts has a new job as a security consultant for a telecom company….

[Owen] Walker pleaded guilty last July — when he was 18 — to a raft of charges connected to his work for an international network that the FBI estimated infiltrated 1.3 million computers and skimmed bank accounts or damaged computer systems to the tune of more than $20 million.

The charges against Walker… were dismissed and he was released without a criminal record after paying a fine and forfeiting cash paid by the criminal group for his expertise.

But after contacting Telstra Clear, the telecom company in question, spokesman Chris Mirams explained that the story was “fairly accurate with the following exceptions”:

“Owen Walker was contracted to be one of three speakers for us at two seminars delivered to customers and prospective customers last October and November. Those audiences included IT, security and senior management. We also used his image for a targeted advertising campaign for our specialist security unit, DMZ Global.”

“He has not presented any seminars to TelstraClear staff, used any computer equipment or had access to our network. He was contracted for those duties only, a period of around two months, and was not, and is not, a fulltime employee…”

“Prior to contracting Owen the company consulted the Police case officer, who was positive in his feedback, and read both the Judge and probation service reports filed with the court. He was, you might remember, not convicted and the Police later publicly stated the outcome was fair.”

The unnamed AP reporter is not the next Stephen Glass, and the main problem with the story seems to be one of semantics and implication rather than facts: “new job” implies full time employment, but does not explicitly state it, and makes it sound like Walker absconded or destroyed $20M. In fact, he was the “ringleader” only so far as he designed the software used in the attack – in short, a botnet author.  In fact, his share of the damage to UPenn’s computer system came to a reasonably low $9526 according to the judge in the case who asked him to pay restitution.

“Black Hat” hackers have gone “White Hat” before – Kevin Mitnick now operates a security consulting company – and similarly to Walker, produced a keynote presentation on computer security called “Art of Deception”, and Kevin Poulsen now writes “Threat Level” and identified 744 registered sexual offenders with MySpace profiles. 

What is different is, perhaps, the methodology – Mitnick and Poulsen, not to romanticize their crimes – operated at a time when hacking was, essentially, a game, and operated primarily alone for the challenge of it – “disorganized crime,” if you will.  On the other hand, Walker used botnets, an attack that only the broadband era would make feasible.  To strain a metaphor, Mitnick and Poulsen targeted individual companies and corporations; while botnets target the general public – the difference between cat burglary and mugging.  Well, mugging a whole bunch of people.

Additionally, the strain that botnets can put on both public Internet and private enterprise networks has placed emphasis on computer security and computer networking working hand in hand – in the field of network behavior analysis.  So… it’s… like mugging a whole bunch of people and making them late for work – okay, the metaphor is falling apart. 

But back to the point at hand – in order to protect the general public from computer-security related problems, like botnets, what we need is level-headed, non-sensational reporting from the mass media (and it doesn’t get any more mass media than the Associated Press.)  Botnets thrive on technical ignorance and misinformation; and it is the role of the press to fight both ignorance and misinformation. 

I just think that the press should be doing a better job here.

This is the penultimate punch line in today’s “Penny-Arcade” Webcomic. It is in reference to a company called “OnLive” which promises to use virtualization and cloud computing to provide broadband, server-side gaming in the browser. The server renders the entire game, then sends back 60 images per second to a browser window. This means, theoretically, that older hardware can “play” complex games.

The big concern, of course, is latency - how many milliseconds does it take from the time you press the button until the button press is registered on the game server, and how many milliseconds does it take for the computer to display the results? In a video interview, OnLive’s CEO, Steve Perlman, handwaved those concerns away:

“If we had a significant amount of lag, it would be unusable. So we had to develop a new technology that would allow the game to run – a lot of the game to run – in the server center. And then really to send tiny pieces of the game down through your DSL and cable modem connection, very very rapidly, with no lag… your computer screen or your TV screen updates so fast that perceptually, it’s as if the game is running right there.”

The video game industry is particularly noteworthy for hype. For example, the Phantom console promised direct game downloads in 2004 – that turned out, of course, to be a scam. And let’s face it – there’s no such thing as “no lag” unless you’re planning to roll out TCP over Quantum Entanglement.

But there’s no doubt that the idea that OnLive proposes is feasible. That is, there’s very little difference between this technology and Cisco Telepresence, only, of course, instead of displaying an image of the opposite partner in the conversation, the image displayed is a computer-simulated hallucination that was pre-rendered and flattened from three dimensions to two. There really is no technical difference between the two.

Yes, it could be done.

But the question is, of course, will something that should work in theory work when it is field tested on the much less predictable conditions of the larger Internet (rather than on corporate LANs or WANs). Ultimately, OnLive relies entirely on network performance to remain feasible, at a time when many broadband providers are purposefully degrading Internet access in one way or another to prevent overcongestion.

Network World linked to a survey conducted by the Internet Society [ISOC].  In it, ISOC claims [PDF] that there are “no concrete business drivers for IPv6” but, paradoxically, customer demand for IPv6 is on the rise. 

Ignoring for a second that “satisfying customer demand” has generally been a good business driver in general, it seems that IPv6 has been, like Duke Nukem Forever, “just around the corner” for quite some time now.  In fact, IPv6 was a Standards Track Protocol back in December 1998, and Duke Nukem Forever has been in development since 1997, which makes interesting betting between friends and technically inclined mobsters over which will happen first: IPv6 mainstream adoption or Duke Nukem Forever releasing.

The survey states in its abstract that:

“While respondents who had begun IPv6 deployment reported gaps in support for IPv6 among tools and applications, they found the process of deploying IPv6 relatively straightforward.”

This is very different from the experiences of the tech guys and gals on Slashdot, whose anecdotes sound a little something like this:

Numbski: “I *tried* to build up a new fiber network in downtown St. Louis using IPv6. I couldn't get the address space!

It's insane - I could get 3x/24 blocks (non-sequential) assigned to my ASN, but in order to get an IPv6 allotment, I had to show proof that I *already* had utilized a full /24 of IPv6 addresses (which is NOT 256. It's 256*256*256!) They said to get it from my upstream provider - they said they don't do that, get it from ARIN. I go back to ARIN, ARIN says "They're full of it, get it from your upstream provider."

Even more insane? IPv6 allotments are FREE! I had to pay per year for an IPv4 allotment, but the free stuff? Pfft...we have it, we'll never run out of it within your lifetime, but you can't have it.”

Slashdot user “Mellon” counters with the idea that providers are probably simply used to the IPv4 way of doing things, which includes fragmentation, huge routing tables, etc., all of which are different from the way things are done in IPv6.

There’s also another thing to consider, and this might make a compelling case for IPv6 – that is, virtualization makes it very easy to create (and to destroy) development platforms in a very, very short amount of time.  With all those machines being moved around, in other words, it’s nice to have a “set it and forget it” IPv6 block that’s large enough to not need to worry about fragmentation.

Finally, the most compelling business case?  It’s simple really.  It’s IPv6.  It’s two more than IPv4.  So if you liked IPv4, IPv6 is IPv4 plus two more!  Seriously, how can you say no to an extra two? 

For those trying to understand the credit crisis as it deals with sub-prime mortgages, there’s a handy video online called “The Crisis of Credit” and explains in simple, easy to understand terms exactly what the heck happened to the sub-prime mortgage market. The problem is that the sub-prime mortgage crisis was only the trigger to the much larger derivatives crisis.

One thing that is clear is that this problem was caused, in part, by a lack of oversight. We all have different opinions on how much, if any, the government should regulate Wall Street. However, when I talk about oversight, I mean that the government didn’t bother collecting the data that they needed to make informed decisions about whether or not regulation was needed, how much, and what type, and as the Daily Show has revealed, the business media didn’t bother collecting the data that they needed to inform individual investors about whether or not particular companies were doing well, or were about to fail.

And just like networking, the simple truth is that you can’t manage what you don’t measure. Forget any ideological ideas you may have about the role of government in business. Even if you believe that the market works best in completely lasses-faire, you can still keep an eye on the market without choosing to interfere in the market.

But back to the crisis at hand – part of the problem wasn’t just that no one was keeping an eye on the situation, but also, that the measurements we did have were just plain wrong. The main reason everyone invested in the top tier of the sub-prime mortgaged based CDOs is because they were rated by the rating agencies as an AAA investment – the same amount of “risk” as treasury bills, the safest investment there is.

But if the insurance companies had the ability to pay off on the insured mortgages, none of us would be in this mess. Instead, here’s what happened – insurers were selling “Naked” credit default swaps. Here’s how Rolling Stone put it:

In a "naked" CDS, neither party actually holds the underlying loan. In other words, Bank B not only sells CDS protection to Bank A for its mortgage on the Pope — it turns around and sells protection to Bank C for the very same mortgage. This could go on ad nauseam: You could have Banks D through Z also betting on Bank A's mortgage. Unlike traditional insurance, Cassano was offering investors an opportunity to bet that someone else's house would burn down…

What this meant is that the insurance companies had to pay off multiple times the value of the original investment if that investment failed.

If this sounds vaguely familiar to you, it’s essentially the same exact plan that Max Bialystock and Leo Bloom hatched in Mel Brooks’ “The Producers.”

I thought about drawing an analogy here to oversubscribing (or overleveraging?) your lines, betting that everyone in the company won’t need to use tons of bandwidth all at once… but when a line gets oversubscribed, the network merely gets congested. Apparently, when overleveraging credit default swaps, the economy blows up.

Of course, if all of this crisis talk is confusing – full of big numbers, strange acronyms, and misunderstood, misapplied concepts – well, there’s a case to be made that Wall Street’s Leo Blooms and Max Bialystocks conduct business in a language that is impossible for outsiders to understand and interpret. (A sort of thieves’ cant, if you will.) Not only has this made it harder for the average person to understand the crisis, it’s made it harder for the media to cover the crisis, and it’s made it harder for anyone not already well versed in Wall Street – in other words, the people who got us into this mess – to sort through all the gibberish in order to get us out of this mess.

I wish that we could claim moral superiority in our own tech world, but I don’t think we can.  Too often, silos persist in IT where protective individuals try to keep their own jobs secure by being the only person who can understand and interpret the data.  It’s the same kind of thing, and I’d say it’s “just as bad,” except that while it may be counter-productive, it hasn’t yet blown up an entire civilization’s economy.

Yet.

In the meantime, here’s some handy links from Amazon:

· Blazer PT-4000 Pencil Butane Torch.

· Speeday Series 60” Pitchfork

And for the kids:

· Angry Mob Playset

Oh, did you know that there’s this basketball something-or-other going on?

Let’s skip to the point. Tomorrow, the NCAA March Madness games start up. For the past few years, the tech media and the NOC teams have been worried about the impact that the games, streamed live over the Internet – and the networks buckled under the streaming recreational traffic.

This year will be the first year in which those streamed games will be Internet broadcast in High Definition.

But to be brutally honest, I don’t think that March Madness will make as big an impact as it did in earlier years – and this has to do with changes in networking philosophies as well as a broader cultural shift.

Six years ago, the first March Madness on Demand hit an Internet completely unused to the idea of streaming video as mainstream. Sure, I remember attempts to sell WWF (now WWE) events via the Internet to college students unable to order them via pay-per-view as far back as 1998, but neither the infrastructure nor the expectation was there; even if we had H.264 encoding back during those halcyon days, we didn’t have the processing power on our client machines to decode it.

But compared to only a couple of years ago, we’re simply used to the idea that you can get video on your computer. There isn’t an HDTV sold in the U.S. that can’t take a computer input via a $30 HDMI cable. When we’ve got nothing better to do, we’re just as likely to YouTube surf than channel surf – and YouTube has videos in 720p – the better to match the 720p videos we’re making with $150 cameras. CBS doesn’t just show March Madness – there’s full episodes of Survivor, The Price Is Right, the Caruso Show (a.k.a. CSI:Miami) – the only show I couldn’t find a full episode of was “Game Show In My Head,” and that’s probably because no one wants to watch “Game Show In My Head.”

ABC, Fox, and NBC also have shows online, not just including Hulu. Great worldly events such as, say, the presidential Inauguration – are also online. And when we think of video on the Internet, we’re not just thinking about entertainment – we’re thinking about medical and enterprise applications made possible by technologies such as Cisco Telepresence. If your network is able to handle all of that – if you’re monitoring and managing your network so that all the other recreational network traffic doesn’t kill it, then March Madness simply isn’t the specter that it once was.

Technologically, our Internet infrastructure is faster, our ability to manage the network is better, and television? Television’s dying. “TV shows” are in name only, and they are “Internet Video Shows.”

So I wouldn’t worry too much about March Madness this year. What damage it could cause was because of its anomalous nature. Now it’s just part of the status quo. If you’re worried about the impact of March Madness, at best, this is a wakeup call. At worst, you’re lagging behind your competitors.

In Greek and Roman mythology, Castor and Pollux were the twin sons of Leda. The God Zeus was the father of Pollux, and the mortal Tyndareus was the father of Castor. They were twins, however. The ancients, apparently, didn’t quite understand genetics like we do.

In the myth, Pollux was immortal, while Castor was mortal. When Castor died, Pollux asked Zeus to let him share his own immortality with his twin.

Fast forward 3000ish years later and we have a similar story of Capex and Opex.

Capex is short for “Capital Expenditure” and Opex is short for “Operating Expenditure.” And at least in an IT context, the two are a lot like the twins, complete with the “mortal” Capex and the “immortal” Opex.

Amazon recently tweaked the terms and pricing of their cloud computing platform, EC2, making it more tempting for CEOs to make “the big switch” to cloud computing.

The cloud – with its operating expenditures – is immortal. That is, you pay continuously, renting computing power, rather than owning it. However, (theoretically) the servers never become obsolete, and maintaining them is the province of an outsourced team, not in-house IT. The traditional datacenter model – with capital expenditures – is mortal. That is, you pay for a server once, and use it until it becomes obsolete.

It can be all too easy to assume that the “correct” decision would be determined by some sort of formula – something like: [(Server Costs)/(Years of use) + overhead + manpower] vs. [Cost of 1 year of Cloud Service]. But it’s a false comparison.

The Cloud and The Datacenter provide solutions to the same problems, but the solutions they provide are slightly different and need to be evaluated for measures other than cost, like, for example, performance.

That is, when moving to the cloud, some performance factors will be completely out of your hands. Connecting through the Internet to the server means your connection to the server is only as good as your Internet connection. Any application which needs to have constant, reliable, low-latency connections will probably be better served in a standard datacenter IT environment – VoIP and VideoIP, in particular.

And of course, there’s the key problem of being at the mercy of another company’s quality control.

Don’t get me wrong – cloud based IT and cloud based software are getting better all the time, and for smaller enterprises they can be cost effective and provide adequate performance. (Note the key word in the previous sentence was “adequate.”) And there are a lot of advantages that cloud computing has over standard data center computing. What cloud computing lacks in performance (and ability for performance monitoring) it can make up for in provisioning – it takes a hell of a lot less time to provision a cloud server through Amazon than it does to order a server from Dell and integrate it into your datacenter. It all depends on what your priorities are.

So whatever happened to Castor and Pollux? Well, Zeus decided to let Castor share Pollux’s immortality, and they were transformed into the constellation we now know as Gemini.

That’s probably a more fitting end to our tale than anything else. It can be hard to play oracle, but if you had to ask me what the future looked like, it would probably look like a combination of traditional datacenter IT and cloud computing working together, each dealing with applications where they’re the best at what they do.

A little less than two weeks ago, Jimmy Ray Purser at Network World posted that Network Management Software “sucks.”

His latest post suggests that there is network management software that doesn’t suck.  As my homeboy Verdi  would say,

Il Jimmy Ray è mobil, qual piuma al vento. Muta d'accento — e di pensiero.

Actually, that’s not fair to Purser - I’m just on a mission to use apply all the things I learned in school to real life at least once before I die, and it was Rigoletto’s turn to become a literary reference.  Anyway, one of the things we wrote in our response to Jimmy Ray’s first article was that: “No single metric is adequate, as [Purser] pointed out, but through a combination of metrics, you can get the data you need.”

Purser’s new blog talks about Cisco’s NAM – or “NAM 2.”

The NAM 2 was the true star of the show. In NMS, many platforms are closed and open ones are pricey. Not so with the NAM. The NAM is an open platform that will work with many NMS software solutions out there today. It accepts input from multiple sources; WAAS Module, SNMP (RMON,hi-cap SMON,ART, DSMON), NDE, etc. I was amazed at the detail I could pull from my network and display it in many different NMS solutions.

Funny how things turn out sometimes. I was looking for a silver bullet in NMS software and it turns out, it was in hardware all along.

Exactly – the idea is to be able to not just grab data but to grab meaning from it by cross-referencing and baselining.  It’s one of the reasons we’ve chosen to closely integrate our products with Cisco NAM – of course, collecting data directly from devices such as Cisco NAM and WAAS not only reduce the data collection footprint but also help us answer one of the toughest networking questions – troubleshooting and monitoring optimized network connections that break the TCP/IP stream.

And don’t forget about all the value you can pull out of Cisco IOS capabilities like IP SLA, CBQoS and NBAR. We have a pre-recorded Webinar on that subject, actually, which you can find here.

It is only through looking at all the data that one can achieve NetZen.  To be truly one with the NetEbb and NetFlow of the NetUniverse. 

I’m also on a mission to use all the bad puns I’ve ever learned, too.

Gmail is currently down for some of the service’s users – and so the world anxiously holds its breath waiting for the resolution.  This follows on the heels of a larger outage two weeks ago.

It should be worth noting that Google’s paid userbase are covered by a 99.9 percent uptime commitment.  It’s not the usual 5 nines associated with most network applications – although in light of recent events…

So, let’s say that we’re talking a total of two days of downtime over the past two years since it was launched to the public on February 7, 2007.  The last outage had Gmail skating on that 99.9% edge, this outage pushes it down to 99.7%. 

But wait, that outage was the third downtime in six months.

Now, don’t get me wrong, Gmail is a wonderful application.  It’s taken the portability of its predecessors and added amazing flexibility – I can get my Gmail email through the Web, on my iPhone, browse headlines through an RSS feed, download to my home Thunderbird client, and even use it as a way to store all those darn game serial numbers in case I lose the sticker that comes on the case. 

Cloud apps provide anywhere, everywhere access, with zero-brain-required “installation,” configuration, and maintenance.  It does a great job of separating content (the stuff you want) from context (how it is presented) so that you can have it presented in a number of formats.  These are all great things, and explains the appeal of the cloud platform. 

For a home user like me, 98% is reliable enough.  But that’s just it – cloud models (currently) can’t provide the same kind of reliability that we’ve come to expect from LAN and WAN applications.  Google didn’t promise “five nines” – they promised three, and delivered one.   I don’t think that it’s a matter of being oversubscribed or of improper maintenance – Google is Google, for crying out loud – but that a cloud application, by definition, contains all the problems associated with any app accessible over a wide area network (in this case, the Internet) as well as the problems associated with serving multiple customers from multiple locations.  The number of mail users a Google or a Yahoo has to support dwarfs the number of users supported at even the largest of enterprises. 

Moving to the Cloud is basically putting your application into a shared resource pool, and it seems to be part of a larger trend.  Virtualization lets us consolidate multiple servers onto one machine.  WAN Optimization lets us consolidate multiple machines into one data center.  What the cloud does is consolidate multiple companies IT departments into one shared pool.  

The point is – when you go to the cloud, you sacrifice a little reliability for greatly increased flexibility.  This is not to discourage you from making a switch over, but just be aware of the risks and, more importantly, be aware of your needs.  In a cloud environment, monitoring network bandwidth remains important because cloud providers will need tools to assess traffic and end-user responsiveness, so that they can adjust their computing capacity to handle the traffic without expensive unnecessary overprovision.

The latest rumors, reported in Techcrunch and other places, imply strongly that Cisco is in talks to buy PureDigital, makers of those little flash-based “Flip” mini-cams. I own three of them myself, but that’s because I like to do things like suction cup them to cars, duct tape them to my helmet while sliding down a 45mph luge, ride with them in a human-sized hamster ball, etc.

The interesting thing about this acquisition is that the Flip camera has greatly simplified the ability for the average user to record and capture high definition video for uploading to YouTube and other sites. Lots of people are doing exactly that – and that’s a lot of bandwidth traveling across the Internet.

Cisco’s interest in Pure Digital may seem a mismatch – Cisco is known as a networking company, where the Flip is a consumer gadget. On the other hand, the Flip is a high-bandwidth gadget – and Cisco can stimulate the demand for its networking hardware and software by stimulating the supply of high-bandwidth applications.

Cisco CEO John Chambers has been aggressively pushing into the consumer space, with some enterprise technologies, such as Telepresence, almost tailor made for the consumer market – assuming you can get the economies of scale to work. Cisco also has a digital media network-attached storage device. The key, it seems, is to get more people using the network and more information on the network in order to feed the need for networking devices. Not so much a “razor and blades” model as a “stubble-growth serum and razor” model.

Of course, Cisco has also bought the consumer-router brand Linksys in 2003, and Scientific Atlanta in 2005. Scientific Atlanta deals mostly in set-tops, cable modems, and digital interactive subscriber systems for VideoIP and VoIP.