Netcosm may look like a video game. But it’s not. It may have explosions and fire, but while it looks like Space Invaders, it isn’t Space Invaders.

Netcosm is an experiment in network data visualization designed to provide a “low cognitive burden” visual metaphor. In other words, it is meant to be as simple to understand as a child’s toy, it is not meant to be a child’s toy.

Still, many people talk about how they’d love to “see the network packets fighting” or would like to see it turned into a game.

I’ll admit that when I first saw Netcosm, it did remind me of the games produced by British game publisher Introversion, which include “Darwinia,” a game about a AI computer simulation, “Defcon,” a global thermonuclear warfare simulator inspired by the 1983 movie “WarGames,” and “Uplink,” about network computer security. All of these games have a “retro” feel to them.

So we’d like to throw this out there: “If you were going to turn Netcosm into a video game, how would you do it?”

DISCLAIMER: This does not mean we’re planning to do such a thing, but it does set the imagination afire. For example, if you could turn network administration into a game you can also turn it into a training simulation. Repeated “playings” of simulated scenarios can help network pros recover faster when the same situations happen in real life. And of course, it would be a hoot if tomorrow’s network administrators and engineers became familiar with the tools of enterprise networking through games played as teens – while people today have to start working for an enterprise before they have any “hands on” experience with enterprise networking.

We’d love to get your thoughts on this.

While we had the chance, we also wanted to thank TechCrunch, Kotaku, and Fark.com for linking to us. The YouTube video has almost 49,000 views, and over 6,000 unique visitors have come to the NetQoS site in the past week.

Silicon.com: Know Your Network

This collection of networking horror stories includes a quote from our CEO, Joel Trammell. They misspelled his name as “Trammel” but that’s okay because they’re British. They spell a lot of words differently in the U.K.

Joel Trammel, CEO of NetQoS, said: "Increasing bandwidth is not a panacea for solving performance problems. Make sure you understand the cause of the problem before taking corrective action like throwing bandwidth at it. Delay for example could be caused by the server, the application or even the transit path. The ability to measure the right performance metrics is key."

Your Comments on Netcosm:

I must admit, it’s a blast seeing some of the comments from places such as YouTube, Tech Crunch, and Fark Tech about Netcosm.

The best comments:

YouTube:

corydlewis: "this is some kinda matrix biznes-go little server packets fight!!!1"

["Go Little Server Packets Fight!!!" is currently the leading candidate for the name of the NetQoS Official Punk Band.]

easybeek: "If at any point in time your data is on fire or exploding, it can't be a good thing."

[Just like purple lightning is never a good sign in a horror movie...]

Fark.com/Tech:

My Liver Hurts: "I'd love to put about 4 flatpanels in my cubicle and get that program running on all of them. Then when people come by to ask me a question or ask me to do something I can just say "[SHUT UP] I'M IN THE MATRIX" really loud and not look at them."

[For real chutzpah, try saying this in a meeting with management.]

MaconL87: "Replace the fireworks in the video with beer kegs and you have an accurate model of how FARK's servers work."

[This is why we're thinking about including a skinning feature...]

dionysos: "I'm going to feel like a real ass if working in a NOC really ends up looking like the one in Hackers, because I've been making fun of that movie for a decade. [but] if the NOC plays 'Voodoo People' in the background, I'm in"

[Voodoo People, Magic People. The who-do-what-you-don't-dare-do-people.]

Techcrunch:

Techcrunch/Jimmydunes: "I won’t be happy until I can monitor my network traffic by smell. “Oh no! Is that burning hair?!"

[When everything works smoothly, the network smells like cinnabuns.]

Our NetQoS Symposium continues into the third day. Here's a rundown of today's events:

First, at 8:00 a.m., there's the registration and breakfast.

Then, at 8:30 a.m. in the Darrell Royal Ballroom, Dr. Cathy Fulton, CTO of NetQoS is going to be presenting our third keynote, entitled "Maintaining Visibility in a World of WAN Optimization."

There is a 15 minute break between all events.

At 9:45 a.m. - 10:45 a.m.:

• "Measuring the Effectiveness of QoS policies" with Shane Bordeau in Darrell Royal A.
• "NetQoS Product Configuration Best Practices" with Stuart Robertson in Darrell Royal B and C.

11:00 a.m. - 12:00 p.m.:

• "Advanced SuperAgent Spanning Best Practices" with Eric Robertson in Darrell Royal A.
• "Maintenance Best Practices" with Stuart Robertson in Darrell Royal B and C.

Those taking the NetFlow Certification Review and Test will need to be in Live Oak from 9:45 a.m. to 12:00 a.m. where Eric Hanson will administer it.

Finally, the Symposium will conclude with lunch at noon in the Hill Country Dining Room.

Netcosm has recieved so much interest and response that people are wondering why we're not offering this as a product or at least offering it as a demo.

The truth is, Netcosm is a cool piece of software developed in our research labs, and we wanted to show people some of the cool technologies we were working on. It was a research product, and we didn't expect the amount of response it got. (Although we admit, we were really, really hoping...)

So if you call up our sales guys and ask for Netcosm, you're going to get the stock response that Netcosm is a research project and is not for sale.

This is for a couple of reasons. While Netcosm is cool, we haven't tested it much in real world situations. Sure, we've got it running on our network but we really aren't prepared to sell a product when we can't guarantee that it will work in our customers networks. It's new code.

Secondly, we may have to recode the thing from the ground-up. Netcosm in its current form is really just a proof-of-concept. However, it requires some hardcore graphics hardware that is more likely to be found in consumer computers. And while the graphics are cool, displaying cubes, spheres, packets, and some simple smoke and fire, should not take nearly as much GPU power as it currently does. we'd like it to take less GPU power. We'll probably We might have to switch out the backend graphic engine - maybe using an early version of DirectX or moving to OpenGL, or maybe even developing our own graphics system in-house.

Because we have so far to go before we can even begin to offer this as a beta version of a product, it would be unfair to customers to promise this to them as a future product.

So, are there any plans to release this as a product? Not at this time - not until we consider it good enough to even consider releasing it as a product.

But don't think the interest has gone unnoticed.

[Editor's note: Corrections made to improve accuracy after receiving clarifications from Dr. Mike Johns.]

Our NetQoS Symposium continues into a second day. Here's a rundown of today's events:

First, at 8:00 a.m., there's the registration and breakfast.

Then, at 8:30 a.m. in the Darrell Royal Ballroom, Issy Ben-Shaul from Cisco is going to be presenting our second keynote, entitled "A Multi-Layered Approach to WAN Optimization and Application Acceleration."

At 9:45, Mike Ellis will present NetQoS Product Vision and Strategy.

We'll resume at 11:00 a.m. and follow this session schedule. There is a 15 minute break between all events, and lunch at noon in the Pavilion outdoors.

11:00 a.m. - 12:00 p.m.:

• "Cisco IP SLA Deep Dive" with Ernie Mikulit from Cisco in Darrell Royal A.
• "Retrospective Network Analysis with GigaStor" with Charles Thompson from Network Instruments in Darrell Royal B and C.
• "NetFlow: Metrics Overview and Traffic Analysis" with Eric Hanson in Live Oak.

1:15 p.m. - 2:15 p.m.:

• "Quantifying Network Performance in an MPLS Environment," with John Mao from Cisco, in Darrell Royal A.
• "Real User Monitoring with Coradiant and NetQoS" with Alistair Croll from Coradiant in Darrell Royal B and C.
• "NetFlow: Security" with Eric Hanson in Live Oak.

2:30 p.m. - 3:30 p.m.:

• "Using SuperAgent to Demystify Application Performance (Part I)," with Tim Smith, in Darrell Royal A.
• "GigaStor Sizing Considerations," with Nathan Bragaw in Darrell Royal B and C.
• "NetFlow Managing Enterprise Applications" with Eric Hanson in Live Oak.

3:45 p.m. - 4:45 p.m.:

• "Using SuperAgent to Demystify Application Performance (Part 2)," with Tim Smith, in Darrell Royal A.
• "Case Study: Deutsche Telekom - Portal Integration of SuperAgent" with Martin Glogowiec from T-Systems in Darrell Royal B and C.
• "NetFlow: User Monitoring" with Eric Hanson in Live Oak.

Finally, at 6:30 p.m, we'll have our Network Social with dinner in the Pavilion with entertainment from the band Reckless Kelly.

What happens when a high profile Web site links to your site?

“Slashdotted.” “The Digg Effect.” Even “Farked.” A product announcement, a cool blog post (on Dungeons and Dragons, perhaps?) and your server gets a sudden spike in traffic.

Ever wonder what that looks like? Sure, you could graph it, but have you ever wondered how it would look like live?

Below, we’ve got a video of the effect (from a simulated “Slashdotting”) and what happens to your network.

Netcosm is a three dimensional video game-like graphical representation of your network and the traffic that traverses it. Initially developed in NetQoS Performance Labs as an experiment in low cognitive burden network monitoring, it allows you to view at a glance the health of the network. Sure, you could get the information from viewing text-based reports and 2D graphs – and this isn’t a replacement for them – but it does allow you to get information about the performance status of your network literally at a glance. Dropped packets are ejected from the stream and detonate in a fiery explosion, servers struggling under load are depicted first as smoking, then as on fire.

Audio cues also determine the health of your network so you don’t even need to be looking at the screen to keep an “eye” on your network – you can “hear” the file crackle and the packets explode, so you can work on other problems while always being aware of the network in the background. Hear the sound of a server crackling or a multitude of explosions?

It’s our way of making hard-to-detect, subtle network performance problems… a bit less subtle, and therefore, easier to detect.

In Netcosm you can move around and zoom in on particular parts of the network or even one server, and you can zoom out to take a look at your entire network. Controls are familiar to anyone who has ever played a three dimensional computer game like Doom, Quake, or World of Warcraft – and that’s not unintentional. The generation coming into the IT rooms today grew up with video games – and will find this tool familiar.

Below is a video of Netcosm in action. Let us know what you think of it in the comments below.

From Network Performance Daily Editor, Brian Boyko.

Tomorrow officially begins NetQoS's three day annual symposium at Barton Creek Resort. We'll be using this blog to keep you apprised of daily events, to share lessons learned, and provide commentary as events unfold.

As for me, I'll be videotaping Joel Trammell's opening presentation on Tuesday morning from 8:30 - 9:30 am, as well as recording video testimonials from our customers. In between, I'll be on my laptop updating the blog.

This year NetQoS is hosting an impressive lineup of speakers including:

• Dr. Issy Ben-Shaul, CTO of the Application Delivery Business Unit at Cisco, presenting "A Multi-Layered Approach to WAN Optimization and Application Acceleration."
• Dr. Cathy Fulton, Co-Founder and CTO of NetQoS, Inc., presenting "Maintaining Visibility in a world of WAN Optimization."
• Dr. Jim Metzler of Ashton, Metzler & Associates, who will be presenting "Ensuring Successful Application Delivery."
• Joel Trammell, Co-Founder and CEO of NetQoS, Inc., presenting "Putting Performance First."

We'll also have some interesting sessions, including:

• NetQoS NetFlow Foundation Course
• TCP Session Dynamics
• VoIP Deployment Best Practices
• Cisco IOS NetFlow
• Monitoring VoIP with NetQoS Products
• ReporterAnalyzer 7.3: New Features and Best Practices
• SuperAgent 7.1: New Features and Best Practices for Incident Management
• NetQoS Performance Center 3.0: An Integrated Approach to Network Performance Management
• NetQoS NetVoyant 5.0: The Next Generation
• NetQoS Product Vision and Strategy
• Cisco IPSLA: A Deep Dive
• Retrospective Network Analysis with GigaStor
• Quantifying Network Performance in an MPLS Environment
• Real User Monitoring with Coradiant and NetQoS
• Using SuperAgent to Demystify Application Performance
• GigaStor Sizing Considerations
• Case Study: T-System's Portal Integration of NetQoS Products
• Measuring the Effectiveness of QoS Policies
• NetQoS Product Configuration Best Practices
• Advanced SuperAgent Spanning Concepts
• Maintenance Best Practices.

In the meantime, we'll be disabling comment moderation on this blog so your posts should appear as soon as they come up. (This also means that the roughly 100 spam posts I get each week will also go through to the site - I'll try to delete them as quickly as I can but until then they'll be on the blog. Don't look at it as sloppy maintenance. Look at it as the opportunity to get low cost ©!@!!$.)

If any of you want to ask a question about this blog or suggest a topic in person for NetworkPerformanceDaily.com, feel free to look me up so long as I'm not currently behind a camera and recording. I look like my picture, have a nametag, and will probably be fiddling with a black MacBook I'm testing out.

In the meantime, here's a rundown of tomorrow's events:

First, at 8:00 a.m., there's the registration and breakfast.

Then, at 8:30 a.m. in the Darrell Royal Ballroom, NetQoS CEO Joel Trammell will be presenting "Putting Performance First," followed by a 15 minute break.

At 9:45, Jim Metzler from Ashton Metzler and Associates will be presenting our first keynote, "Ensuring Successful Application Delivery," which will identify the current roadblocks to ensuring acceptable application performance and describe a framework that IT organizations can use to mitigate the technology and organizational challenges, and improve their ability to deliver applications with acceptable performance.

We'll resume at 11:00 a.m. and follow this session schedule. There is a 15 minute break between all events, and lunch at noon in the pool area outdoors.

11:00 a.m. - 12:00 p.m.:

• "TCP Session Dynamics," with Kevin Davis in Darrell Royal A.
• "VoIP Deployment Best Practices" with Jeff Hicks in Darrell Royal B and C.
• "NetFlow Overview and Background" with Eric Hanson in Live Oak.

1:15 p.m. - 2:15 p.m.:

• "Cisco IOS Netflow," with Jean-Charles Griviaud from Cisco, in Darrell Royal A.
• "Monitoring VoIP with NetQoS Products" with Jim McQuaid in Darrell Royal B and C.
• "Network Monitoring, Measurement, and Management" with Eric Hanson in Live Oak.

2:30 p.m. - 3:30 p.m.:

• "Reporter Analyzer 7.3: New Features and Best Practices," with John Mao, in Darrell Royal A.
• "SuperAgent 7.1: New Features and Best Practices for Incident Management" with Tim Smith in Darrell Royal B and C.
• "NetFlow Packet Archetecture" with Eric Hanson in Live Oak.

3:45 p.m. - 4:45 p.m.:

• "NetQoS Performance Center 3.0: An Integrated Approach to Network Performance Management," with Ben Erwin, in Darrell Royal A.
• "NetVoyant 5.0: The Next Generation" with Jeff Hicks in Darrell Royal B and C.
• "Types of NetFlow and Deployment" with Eric Hanson in Live Oak.

Finally, at 6:30 p.m, we'll have our Network Social with dinner in the Wildflower Atrium followed by Casino Night at Live Oak.

By Brian Boyko

Until recently, I underestimated the value of a good error message.

Recently, I wrote two articles for HardOCP detailing my experiences with Ubuntu Linux and with Windows Vista. Both have flaws and high points, and I'm not going to get into playing the "which OS is better" game.

But there is one area, however, which I do want to compare, because it caused me professional consternation and it is one that is usually overlooked.

Error messages.

(Continued…)

Continue reading "Editorial: The value of a good error message." »

First, an apology to the readers. Usually I try to add some entertaining, humorous commentary that make you want to read more, brightens up your day and sometimes makes the oftentimes dry world of IT news a little easier to swallow.

But today we've got some comments on IT management from Network World, an instructional video on Amazon's outsourced virtual servers from ZDNet, some talk about Verizon's EVDO network from wired, and some developments in AI from the University of Illinois at Chicago.

In fact, if you have something I can work with, let me know about it in our comments section or by e-mailing brian.boyko@netqos.com.

Anyway, on with Thursday's links…

ZDNet: Time to throw away your servers? (Video)

ZDNet's executive editor, David Berlind, talks about Amazon's Elastic Computer Cloud, which is a pay-as-you-go data center outsourcing service that uses virtual servers. The advantage of those virtual servers is partially that you can turn off the servers you aren't using during less than peak hours, and that the price for using one of these virtual servers is less than the price of outsourcing a physical server.

Network World: IT Management Done Right:

Based on interviews conducted by Network World editor Bob Brown at Babson College, this article helps explain how to run an IT department while integrating newer technologies such as VoIP, Virtualization and WiFi into the existing architecture.

The QoS that I find relevant is not at the bits and bytes level, it is all about the "perceived performance" from the users' perspective. This is not to say we take all user comments of "the network/application is slow" at face value, but we do try to find patterns in these comments and investigate the causes that lead to the comments. If there is one standard measurement in computing, it is that happy users don't complain. Therefore, we take all complaints seriously. Setting arbitrary performance levels at the network layer is not useful to me, the entire experience of the user has to be taken into account. This usually means we look at the application on down when researching performance/quality of service issues.

In short, the line between the network vs. servers vs. application is so fuzzy you have to take a macro view toward QoS, no matter what your job role is. I expect everyone within my organization to look at the user experience first and then their statistics to support/refute the claim.

Wired: Verizon says Unlimited equals 5GB: Bans P2P

Wired Magazine's GadgetLab blog talks about some of the fine print in the EVDO network from Verizon, which prohibits use of P2P and streaming media, and limits transfers to 5GB. The main thrust of the article is that they're marketing this limited functionality as "unlimited" - but I'm more interested in the idea that Verizon is likely doing everything to cut down on traffic on the EVDO network.

"A person engaged in prohibited uses, continuously for one hour, could typically use 100 to 200 MBs, or, if engaged in prohibited uses for 10 hours a day, 7 days a week, could use more than 5 GBs in a month [...] Anyone using more than 5 GB per line in a given month is presumed to be using the service in a manner prohibited above, and we reserve the right to immediately terminate the service of any such person without notice."

Sounds like poor capacity planning leading to poor customer relations to me.

Uinversity of Illinois at Chicago: Virtual Reality for Virtual Eternity.

Virtual Necromancy.

Imagine having a discussion with Isaac Newton or Albert Einstein on the nature of the universe, where their 3-D, life-sized representations looked you in the eye, examined your body language, considered voice nuances and phraseology of your questions, then answered you in a way that is so real you would swear the images were alive.

It's interesting stuff. I'm curious to see if this could practically be sent over the network, however, and what the bandwidth requirements would be.