Cisco Live! a.k.a. “Cisco Networkers” is going to start the week after next on June 27^th, and I thought it might be time to start plugging our appearance at the—wait a minute…

“Cisco Systems.”

Based in San Francisco.

Oh my god.  I just got that.

I feel like such an idiot now.  Really – I thought it was named after someone like a John Cisco or something.  Like he was the founder of the company.  I’m probably thinking of John Chambers, the current CEO, but the company was founded by Len Bosack and Sandy Lerner.

What else am I missing out on?

Anyway, NetQoS, as “diamond sponsor” of the event, we’ll will have a “Super Session” on Tuesday, June 30, at booth 807 at the Moscone Convention Center in San Francisco, and--

I’m not going to be able to get that out of my head now.  It’s like seeing the arrow in the Fedex logo.  What else have I been missing in my life that’s so obvious to the people around me?  I feel like Grover, who found out that the monster at the end of the book was him all along.  (Him and his drinking problem.)

I’m going to need some time to think about this. 

Well, coming from the land of the ice and snow, from the midnight sun where the hot springs blow, the Norwegian based Opera Software is following in the footsteps of the Vikings.  Opera Unite is a technical achievement and if it – or a worthy imitator - is widely adopted, will be a game changer.  The Vikings were the game changers of their day, engineering the best seafaring technology of the time.  They also pillaged and burned a whole mess of Europe, which many people may find reprehensible, but, hey, the Vikings were the best at what they did. 

At its core, Opera Unite is nothing more than a webserver – we’ve had these things for years, of course.  What makes it different is the ease at which it can be set up – download the browser, create an Opera.com account, and you’re done.  By routing through Opera’s servers, you don’t need to mess with VPN, Remote Desktop, IP addresses, or configuring those fiddly little port forwarding settings on your home router – you don’t even need to have access to it – or even to know what a router is – to use Opera Unite.

Right now, the immediate use is to run it on the home computer to access files from work, and vice-versa.  And that’s where the snag comes in.  Opera Unite cannot be considered in any way “secure” – the fact that it connects to a third-party server makes it ripe for a man-in-the-middle attack, people might mistakenly share sensitive information on their work or home computers, and of course, there’s the problems that you’d expect to have with any filesharing app on the PC.

But more importantly, this will have a major impact on performance.  Employees running data servers can choke the network links your company pays for; we have already seen this when highly technical users run FTP or Web servers from their office desktop machines.  Opera Unite doesn’t change the nature of FTP servers – it just places creation and access of FTP servers and hosting services in the hands of the many, rather than the few. 

Even so, I can’t help but think that overall, Opera Unite is a good thing overall.  It reduces dependence on third-party hosting sites and cloud apps – like YouTube.  If there’s a video that YouTube doesn’t like – or has to take down because of a mistaken or fraudulent copyright claim, the video can still be made available.  (Often times, copyright law is abused to get hosting services to remove unflattering footage of a company or organization.)  And if a particular hosting service should die, the data can still be accessed.  In this manner, it removes some of the risk from cloud computing by allowing anyone to run their own hosting services. 

It doesn’t hurt Opera that they’ve found a hail-mary pass for desktop software to remain relevant in the age of the “software as utility” philosophy of cloud computing… by essentially providing a desktop app that turns your desktop into a cloud computing platform. 

Additionally, although I don’t hold out much hope for it, it may increase demand for more uploading capacity for home users.  AT&T DSL, for example, maxes out at 768kbps upload speed – over three hours to upload a single gigabyte of information. There’s been little complaint about this because most people care more about download speeds than upload speeds – with their own hosting services however, people might be more likely to notice, and care, about what speed they can access their home computer from work, or their friends or colleagues can access their files.  That may lead to increased upload capacity provisioning. 

Ultimately, I’m psyched about Opera Unite as a desktop user, but in the trenches of IT, I’d want to make sure I had a way to track this traffic and see if it affects network performance in a meaningful way.  It may be nothing but a flash in the pan, but if it does catch on, I’d rather err on the side of caution.

Recently, Chandra Hosek, who maintains the NetQoSLive Twitter account located to the left of this blog-post, unless you’re reading it sometime in the far future after which we may have redesigned the site meaning it could be anywhere on the page, or even replaced by a service even more twitter-like than Twitter, or perhaps you’re reading this from a space station and you’re a robot.  A robot that can feel love, and question its existentia—

Where was I going with this?

Oh, right.  (That’ll teach me to get 6 hours of sleep over 48 hours.)

Well, recently, Chandra started asking people what some of their favorite network engineering blogs were, so that we could not only add them to our blogroll on the site, but maybe there’s a blog out there that’s really got some relevant information and we’ve missed it. 

I know – obviously NetworkPerformanceDaily.com is your “favorite.”  But we’re looking for vendor neutral blogs, specifically, so NPD wouldn’t qualify.

Right now we’ve got a couple – there’s The Network Zone, which has a great post up on how to put value in your network where it matters most, and Should Have Gone With Cisco, which recently put out a walkthrough on how to configure the Cisco Nexus 7000 router virtual portchannel feature – but we could use more.  Could you shoot us some links in the comments section of this post?  Thanks!  

There have been many jokes about the irrelevance of Twitter – or at least about the irrelevance of much of the content on Twitter. Some of them have been amazingly creative and funny, like the Tonight Show’s “Twitter Tracker,” and some of them not funny at all, like any of the jokes we’ve made on this blog on the same subject.

But the events of the past few days have killed Twitter irony. In a country where “subversive” blogging is punishable by death, Twitter (along with YouTube) has been the go-to to get information about the happenings in the tightly-controlled Iran.

Furthermore, the relevance of the newest of the new media highlights some of the irrelevance of the “traditional” media. The Revolution, it seems, will not be televised – CNN.com didn’t mention the unrest for days. As ReadWriteWeb.com put it:

“Hours after Iranian police began clashing with tens of thousands of people in the street, the top story on CNN.com remains peoples' confusion about the switch from analog TV signals.”

CNN, the TV Station, provided only regular news reports instead of wall-to-wall coverage akin to the coverage of the Tiananmen Square protests or the first Gulf War – the events which practically made the CNN news channel. If 24 hour news won’t provide 24 continuous hours of news coverage on the most important subjects, what, pray tell, is the point of 24 hour news stations?

Instead of covering Tehran, CNN showed a rerun of the Larry King show, where King interviewed the stars of “American Chopper.” Now, I get it, sometimes newsdays are slow, and sometimes you need to fill in the gaps. But this was anything but a slow news day.

Now, you could make the argument that Twitter is reporting, among other things, rumors and mistakes. If the information coming out of Iran is accurate, CNN just got shown up by a better news service; if it’s inaccurate, CNN should have been dispelling those rumors.

In contrast, Twitter (the company and service), recognizing the importance of the news from Iran, delayed it’s scheduled downtime so that it can remain available for those Iranians who can still access it.

In the meantime, Iran’s government is doing everything they can to prevent the news from getting out – a futile effort in most cases. Looking at the Iranian Internet services, you can a clear pattern of additional outages and unstable connections – starting on Saturday. (Kudos to Renesys to making this information available.)

Of course this is encouraging for encouraging the promise of Democracy. For good or ill – in this case, good – it’s extremely hard to fully block comment and communication on the Internet. There are still sysadmins out there who think that blocking YouTube is an effective response to over-subscribed enterprise networks. Iran, a dictatorship, with an army and a nuclear program for crying out loud, can’t block YouTube completely – what makes you think you can? This effectively illustrates why the appropriate response is shunting YouTube traffic into a lower quality of service so that it doesn’t interfere with business applications, and running network traffic analysis to make sure that it doesn’t.

Iran’s governmental system is interesting because unlike many other world dictatorships, it sets up an expectation of democracy; the idea being that concepts of voting, parliament, democratic representation – they’re not only not foreign to the Iranian culture, but, as we can clearly see from the protests, Iran has one of the most – if not the most – vibrant democratic cultures in the Middle East, in complete contrast to having one of the least democratic governments in the world. Which makes the YouTube and Twitter coverage extremely important - what many Americans are learning from it is that Iran is not a country of extremists and radicals, but a modern, progressive nation with a repressive, barbaric government.

I wish ‘em luck.

When Savvis promises “proximity hosting,” they mean it – according to this New York Times Magazine article. In Weehawken, New Jersey, right outside of the Lincoln Tunnel, there’s a data center that houses the Philadelphia Stock Exchange’s computers. (The PSE is now part of Nasdaq.) Firms compete to have their computers located close – physically and in the networking sense – to the trading exchanges in that data center. Milliseconds of latency are unacceptable in this environment.

“It used to be that things were done in seconds, then milliseconds,” Varghese Thomas, Savvis’s vice president of financial markets, told me. Intervening steps — going through a consolidated ticker vendor like Thomson Reuters— added 150 to 500 milliseconds to the time it takes for information to be exchanged. “These firms said, ‘I can eliminate that latency much further by connecting to the exchanges directly,’ ” Thomas explained. Firms initially linked from their own centers, but that added precious fractions of milliseconds. So they moved into the data center itself. “If you’re in the facility, you’re eliminating that wire.” The specter of infinitesimal delay is why, when the Philadelphia Stock Exchange, the nation’s oldest, upgraded its trading platform in 2006, it decided to locate the bulk of its trading engines 80 miles — and three milliseconds — from Philadelphia, and into NJ2 [in Weehawken, NJ], where, as Thomas notes, the time to communicate between servers is down to a millionth of a second. (Latency concerns are not limited to Wall Street; it is estimated that a 100-millisecond delay reduces Amazon’s sales by 1 percent.)

Back in March 2008, electronic trading made up 60-70 percent of the daily volume of the NYSE. (I’m sorry I don’t have more recent numbers, but they might have been artificially affected by the credit crisis anyway.) And when you remove human beings from trades; the only thing that matters is the speed of a sale; whichever seller’s computer connects first to the buyer makes the sale, whichever buyer connects to the low-bidding seller first gets the bargain. Speed, while not everything, is not underestimated – and it’s one of the reasons you need to identify immediately any problems with network performance in financial applications. Every second a problem doesn’t get fixed – even problems that are imperceptable to the end user, like an added 3ms of delay - means more money is lost.

Now, if your company is over-leveraged and built on shaky investments, network performance won’t save you – we’ve seen a lot of companies with very good network infrastructures go downhill these past few months.

If you want to learn more about the topic of monitoring trading applications for performance, you might want to check out Alex Malone, Software Engineer Manager at NetQoS, who will be speaking at the Securities Industry and Financial Markets Association Technology Management Conference & Exhibit on June 23-25 in NYC. Alex is scheduled to speak June 24, at 2:35pm. You can also look us up at booth #1822.

I got some good news today. Hulu.com is putting the first season of my favorite TV show, ever, “The Mole,” online. (Seriously. Watch it. It’s good.) There is no doubt that video on the Internet has left the realm of novelty and has entered the mainstream; many people watch as much or more “tv” delivered via the Internet then people actually watch a physical television. Some, like myself, only watch video downloaded via the Internet.

The upshot of all of this is that Cisco recently updated their Visual Networking Index which tries to forecast where the Internet will be in the future. And it’s breathtaking.

• In 2008, the Internet transferred 9 exabytes (1 exabyte = 1.074 billion gigabytes) worth of data each month. Cisco predicts it will carry 56 exabytes per month by 2013. This means that annual traffic will reach around 2/3rds of a zettabyte (1 zettabyte = 1024 exabytes).


• The boogieman of years past, P2P traffic, will continue to grow in absolute terms but decline as a percentage of Internet traffic. Internet video, now 1/3^rd of all consumer Internet traffic not including P2P file sharing, will account for over 60% of all consumer Internet traffic by 2013. Including P2P, TV, and Video on Demand, this will account for over 91% of all consumer traffic by 2013. Video communications traffic growth is small but accelerating, real-time video is growing in important, and VoD traffic will double every two years through 2013.


• Mobile phone data traffic will also double every year through 2013, and 64% of that will be video data – which gives a lie to the idea of calling it a mobile ‘phone.’


• While today, business Internet traffic is twice the volume of WAN traffic, but video on the WAN will result in business WAN traffic growing at a faster pace than business Internet by 2013.


• Twitter, Youtube, and Facebook will merge to a single social networking site, called “YouTwitFace!” – No, wait, that one was from the Tonight Show with Conan O’Brien.

One of the traditional solutions to solving online video problems in Enterprise environments has been to block all Internet video entirely; as businesses will require more video information to collaborate between other businesses and with customers; this becomes a less viable option. The trick is managing network video to make sure it co-exists with your data applications – and the sooner this happens, the easier it will be in the long run.

Earlier, we wrote about how March Madness affects Internet performance every… well, March, and as such, it was such a predicable event that IT pros should be aware of the effects of the tournament by now. This growth in video is no different, it’s just constant – a March Madness every day.

This means that if you haven’t already prepared for network service in the era of Net Video, it’s time to step up, think about whether or not to filter packets, what to filter, how to protect your data applications, know which applications need which quality of service, and understand the differences between quality of service tiers. Five years from now, handling Internet video will be the main priority of the network engineer’s job; so it’s a good idea to study up today.

A quick summary of theAberdeen Group Benchmark report, “Getting IT on the C-Level’s Agenda,”:

It ain’t.

Well, it’s not on the agenda of most companies and Ann Bednarz at Network World explains how bad the situation is – despite years of hammering home the importance of identifying problems before users are effected, 61% of the respondants in the Aberdeen study said that they couldn’t do that. Despite the importance of measuring the business impact of application performance, 49% said they couldn’t do that. A full 46% even said that they lacked visibility into the end-user experience.

And the companies affected by these problems cited decreases in employee productivity and customer satisfaction. Conversely, those companies that took steps to address these problems reported an 83% success rate in preventing issues with application performance before users are impacted, with 113% average improvements in application response times.

That companies have problems in these areas isn’t unusual; but the percentage of companies that haven’t addressed these problems this late in the IT game astounds me. For example, take this quote from the Network World article:

“Originally we thought that we didn’t have enough bandwidth to run an ERP application and were thinking about adding an additional T-1 line. However, a technology solution that we had in place helped us realize the root-cause of the problem was on the application side -- not the network,” the IT director shared with Aberdeen. “This allowed us to avoid adding more bandwidth, but it also allowed us to be able to effectively manage application performance on an ongoing basis and address potential problems before they could impact end-users.”

If there’s one thing that we’ve been trying to hammer home with this blog, (other than that D&D is awesome and that you shouldn’t knock it until you try it,) it’s the idea that without visibility you’re not going to be sure where network problems lie, and you’ll end up wasting money on guesswork solutions which may not actually solve the problem in the first place.

More importantly; having visibility into your network and application performance allows you to make a very shrewd business decision. Getting “the best possible performance” out of your network is good, but in reality, most often enterprise IT is tasked with finding a baseline acceptable performance, and finding the lowest possible price that it takes to achieve that performance. Of course, you can’t tell what acceptable performance is unless you know what “normal” baseline performance is, and you won’t be able to tell what’s acceptable unless you have visibility into the end-user experience.

For a politics junkie and a technology junkie like myself, the EU Elections produced results that were really, really interesting to watch.

First off, how about the wacky U.K.? Labour, which currently is the dominant party in the U.K. Parliament, wins only 15.7% of the vote – a third place finish behind the Conservatives and the Euroskeptic UKIP. The UKIP, which seeks withdrawal from the EU, only got 2.2% of the U.K. national vote in 2005, though it’s tough to figure out if the radical change over the past four years is due to the recent expense scandals, the EU having a proportional representation election system while the U.K. uses “winner-take-all” voting, a great sea change in the voting populations’ opinions, the idea that a Euroskeptic party would naturally receive more prominence in an EU election than a U.K. one, or some combination of the above.

Some days, I wish we had a NetVoyant for voters, so that we know what’s going on when they vote. I suppose we’ll have to make do with the more traditional type of polling for now.

Another interesting election result from the U.K. was the establishment of the BNP – the British National Party – getting 6.2% of the vote, putting them in 6^th place overall, and entitling them to two seats. If you are not familiar with the BNP, they are, without exaggeration, the Nazis. That’s not a joke - the BNP’s political party platforms are expressly based on racism (including but not limited to anti-Semitism) and fascism. In fact, the BNP is so universally loathed that membership in the party is grounds for dismissal from serving as police officers, as the BNP is “totally incompatible with the duties and values” of police forces.

6.2%. Damn. Where’s Indiana Jones when you need him?

Well, at least there’s some good news; the Pirate Party of Sweden (PiratePartiet) won its first seat in the European Parrrrrrliament, and came close to a second one with 7.1% of the Swedish vote. This is huge. The Pirate Party did not exist four years ago – now it has a MEP (Member of European Parliament). Even if you disagree with the Pirate Party’s stances, it will bring more prominence to technology issues which have too often (for us techies, anyway) been relegated to the backburner of political concern.

Representation means more coverage of technical issues in the mainstream press, as well, which will hopefully cut down on the number of conversations you have with people asking you “What is this BitTorrent thing,” or “Why can’t you have ‘Born to Run’ playing in the background of the corporate Web site if Springsteen himself put it on YouTube?”

PiratePartiet also claims that they received more votes than any other party among under-30 year olds; which, if true, may make this a generational sea change.

Or it could just be a flash-in-the-pan, who knows.

On Tuesday of next week, Microsoft plans to unleash a large patch – the largest “Patch Tuesday” in eight months. If you update your desktop systems individually, you might want to prepare for that traffic.

In many ways, the smaller size of patches seemed, at the very least, to imply that the codebase behind the Windows OS and Office might have been stabilizing, or at the very least, that the XP and Vista codebases were wrapping up in anticipation of the Next Big Thing, Windows 7.  Apparently, that’s not the case.

When I was a younger, brasher, more hot-headed geek in college, the very name Microsoft would conjure up images of hatred.  When Windows XP came out, I had heard rumors of something called “Palladium” and draconian DRM measures – and, let’s be fair here – this was around the time Microsoft’s competitors were starting to get good.  Despite competition from MacOSX and Linux – very good competition, I might add, - Windows is still the OS of the masses, Office has not been displaced as office king-of-the-hill since it displaced WordPerfect, despite attempts by Sun, OpenOffice.org, Zoho, and Google.  (That may change in the future, but not likely, according to Forrester.)

But it’s been an interesting road for Microsoft; XP SP2 wasn’t released with a whole lot of fanfare, but its basic competence (well, it compared favorably to previous versions of Microsoft OSes in the stability area, and similarly to contemporary MacOSX 10.2.)

Then Vista came out, and suddenly the old hatred for Microsoft is horrible again.  To be fair, I trashed Vista in a review for HardOCP when it first came out, as it seemed like nothing worked; most of these problems were fixed with the service pack, but it was still more problematic than Windows XP while still doing the same basic job; this is one of the reasons that business adoption for Vista remains – well, let’s just say that most businesses will probably skip Vista and head straight to Windows 7. 

Which brings us back to Patch Tuesday.  If it wasn’t apparent before, it seems that no operating system is “done” anymore, like the Windows 95 and 98 days; (Even 98 had “98 second edition”) and that even the most basic part of the computer – the operating system – requires the network.  But it also implies that patching is not something we’re ever going to grow out of – that operating systems are never done.  The open-source model, of course, relies on “release early, release often” but it seems that through Windows 7, Windows 8, Windows 9, and Windows 10 (and Mac 10.6, 10.7, & 10.8) we’ll be waiting for that big glob of data to download every week, to keep our systems up to date.

On the other hand, Patch Tuesday is repeatable, predictable – hell, it’s so predictable, we call it “Patch Tuesday.”  It’s a great way to determine whether or not your network can handle a spike in traffic, and how well it performs under those conditions. 

GM and Chrysler, combined, will be closing about 3000 dealerships across the U.S., leaving many out of work, and hurting local economies across the country.

On the bright side, however, this is a strange opportunity for networking companies. See, car dealerships and the car companies they franchised with, were some of the earliest adopters of electronic data interchange (EDI) networks; the idea being to connect a large number of automobile dealers and lenders. The EDI systems provided an alternative to paper documents, meetings, faxes, etc., which is one of the major reasons that you can walk into a dealership today and drive off the lot with a new car, instead of waiting a few days for the paperwork to clear. But it also meant that car companies could have dealers select cars to order, and have the network apps automatically order parts for building the car from suppliers, issue instructions on how to assemble the car, invoice the dealer, and pay suppliers. Enterprise IT networks, therefore, are a major priority for many if not most dealerships.

So my question is – what’s going to happen to all the networks running to car dealerships? Some dealerships will rebrand, in which case they’ll probably keep the capacity but route it to the new affiliate, but others, who either close up shop, sell to a non-dealership, or go the used car route. This seems to me to be an interesting, under-tapped networking resource.

Because – and let’s be frank here – there are an awful lot of car dealerships in America, and 3000 car dealerships is 60 dealerships per state. Granted, it’s highly unlikely that the closing car dealerships will be uniformly distributed geographically, but anyone looking at starting up municipal broadband – or even a for-profit Mom & Pop broadband shop - might want to take a look and see if there’s a closing car dealership in the area.