We ran into Luke Kanies, the founder of Reductive Labs, which develops the open source server and desktop automation tool "Puppet."

Puppet is designed to let you run system administrative tasks automatically on multiple systems – even if those systems use different operating systems. All you do is input the configuration you’re used to handling, and Puppet automatically configures the servers for you.

Below, Kanies talks about Puppet and open-source development for enterprise and system administration projects.

By Brian Boyko
Editor, Network Performance Daily

There's an essay that's going around some of the top news sites, called "If Wishes were iPhones, then beggars would call," about the idea of 3rd party tools on the iPhone.

The main point of the article is this:

Apple has been unwaveringly clear that the iPhone is theirs. Not yours, not Ambrosia's, not J. Random Hacker's. You may own the hardware, but you only have a limited license to use the software, and an ongoing contract to use the network. If you don't like those terms, your only recourse is to shop somewhere else to begin with….

I don't understand this continuing obsession with buying things that you need to break before they do what you want.

I don't know what it is either; I just know that it exists. It is the urge in all mad scientists to go grave robbing for old, discarded technology that can be put to new use, to twist and mangle things into doing what they were never designed - or ought never - to do, to cackle with glee as we defy the laws of man, God, and nature to raise our voice up to the heavens, and scream against the thunder, "LIFE! GIVE MY CREATION LIFE!"

My first computer was built in 1996 - back in the days when IRQ conflicts were a serious problem and required setting jumpers to modify - by a family friend from old discarded PC parts that would have otherwise ended up on a rubbish heap. We dubbed it "Frankenstein."

The end result of all this stuff is an entire geek subculture, only starting to show through - of hardware hackers. In two weeks, for example, the Maker Faire is coming to our home town of Austin where people make stuff by breaking stuff - there's an entire sub-exhibit on taking old children's toys and making them into strange instruments - called Circuit Bending. There's a few displays which show off the capabilities of the Arduino, an electronics prototyping board which can be configured for a variety of purposes. And on the networking side; demonstrations of how to build your own Beowulf clusters.

But that's the geek mentality. Create. Invent. Remix.

Apple's whole business model is stifling those tendencies in exchange for providing a simpler product - and that's not a bad thing for Apple, because they get to sell to the 90% of the human population that aren't geeks and don't want to bother with figuring out new and exciting ways to do things. They even stopped carrying a mid-range tower line of computers. But if you notice, Apple sticks well to the consumer side of the market, not the enterprise side - sure, they have the Xserve, but they really haven't followed up with other entries into the enterprise market since its introduction.

In enterprises we've seen a trend over this past decade from big, proprietary system management frameworks to individual tools that work together, and I think that can be attributed to the Frankenstien Syndrome. Even Cisco is breaking up their IOS into different modules. We here at NetQoS sell all our products individually, even though they all work together through the NetQoS Performance Center Web portal. We even made it a clear goal to allow third-party Web applications - any third party Web application - to work with the NetQoS Performance Center. Figure out a way to map network slowdowns to Google Maps? Go ahead, we won't stop you.

Because it's important to harness - not stifle - those geek tendencies which make geeks so well suited for enterprise networking.

By Brian Boyko
Editor, Network Performance Daily

I'm supposed to be a technogeek, right? Up on the latest and greatest gadgetry. I get called on international radio morning shows to talk about the latest trends, I'm a professional blogger - a career that didn't exist half a decade ago, I'm a smart dude with all the latest tech toys, doing viral videos with my $100 flash-memory vidcam. I'm 100% Web 2.0'n, YouTubein', Facebook Friendin', RSS Feedin', Linuxin', Twenty-First-Century-Digital-Boy.

But my cellphone is about four years old and I don't use it for anything other than making phone calls. No texting, no Web browsing. It just makes phone calls. And I don't plan on upgrading anytime soon.

My friends - especially those who live overseas - all wonder why I don't yet have an iPhone, the "wundertoy" of the century. The answer can be found in Cory Doctorow's post for Information Week. In short, the iPhone sucks on purpose. Sure, it's slick looking, has a beautiful interface, and provides a great playlist. But, does it live up to its full potential? Hardly. And you have vendor lock-in to blame.

It's possible to use a song in your iTunes library as a ring tone, but someone at Apple decided they wouldn't let you. It's possible to use the iPhone on multiple carriers, or to not lock-in customers for two-year service agreements, but someone at Apple decided, for whatever reason, they wouldn't let you. It's possible to switch out SIM cards so that you'd be able to use the iPhone internationally, but - well, you get the pattern by now.

This is where VoIP comes in.

Already, we have pocket-sized computers that we carry with us - Palm Pilots and Blackberries. Despite the lukewarm acceptance of Microsoft's "Origami" initiative, ultra-mobile computing isn't that far off.
Desktop computing took off because it was a general purpose device. You can play games with it. You can send messages with it. You can write papers with it, you can draw, you can compose music - you can do all those things with the right software on general hardware. That's the big draw.

So when the iPhone comes out, for all intents and purposes a semi-powerful computer that you can hold in your pocket, the first thing Apple does is lock it down so that you can only do what Apple says you can do. It turns a general purpose device into a limited purpose device.

Additionally, despite setbacks in getting municipal WiFi off the ground, I can't see, for the life of me, the possibility that wireless access points will decrease over the next decade - only that they're not increasing at the rate that we'd like. It won't be long before Wi-Fi coverage overshoots many of the major phone carriers cellular networks - it has to, because the cellphone companies have chosen non-interoperable standards and done their best to lock people into them. Wi-Fi, on the other hand, generally pushes out the same bits, no matter where you are in the world. So while AT&T may have deep pockets, the larger number of smaller pockets putting out Wi-Fi spots around the world will move faster and change with the technology instead of trying to halt technological change in order to milk more money out of an obsolete business model.

Companies such as Skype have already laid the foundation for VoIP's replacement of the cellphone network. Companies such as OpenMoko are working on the hardware aspect of cellphone computing. I'd be surprised if these two industries weren't already talking to each other about creating a "chocolate and peanut butter" solution which will make VoIP over WiFi the killer app of the cellphone, and cellphone connectivity the killer app of VoIP over WiFi.

It's an idea that Cisco had a while back. Remember the brouhaha over the "iPhone" trademark? Well, it turns out that Cisco put out the "Linksys iPhone" a while back - a phone designed to make calls over Skype. At $120 from Amazon - with only Skype fees to pay after that, this is a technology which fills a need. The only problems are that it is currently overshadowed by the Apple product of the same name, and that, for right now, the cellphone companies do have better network coverage. That won't last long though.

The big elephant in the room is that no one really likes the cellphone companies. Indeed, a great part of the AT&T/Apple deal was Apple banking on the goodwill of the Apple corporate brand name. And so all it takes to upset the apple cart of their market share is for someone to offer a product that offers everything - including the same or better call quality - that the cellphone companies do, and not "suck on purpose." (Oh, and maybe a slick looking slim-line case with chrome accents.) When that happens, VoIP is going to take off not just for enterprise networking but for personal computing - and VoIP monitoring becomes that much more important for providers to ensure call quality, and network administrators to make sure that VoIP traffic doesn't interfere with business.

By Brian Boyko
Editor, Network Performance Daily

Slashdot recently linked to an article from MacWorld showing that the amount of time that people spend watching online video has steadily increased. (In other news, water is wet…) Google's YouTube and Google Video served up over a quarter of all internet videos.

I think we can assume that a fair percentage of them were watched from corporate networks. Not just because of recreational use but because video is a very compelling medium that can convey work-related information, sometimes more quickly and more accurately than text alone.

For example, our Whiteboard Series was created with the expectation that people would watch our videos on WAN Optimization and VoIP from work, where they would find the information most useful.

One really can't just block YouTube, or Veoh, or Yahoo Video and expect blocking it to solve the problem of tying up vital bandwidth, because video is increasing not just as a 'bandwidth hazard' but as a method of communication. And it's only going to get more bandwidth heavy - and more useful - as MPEG4+ACC "Moviestar" Flash Video, or WMV using Microsoft's Silverlight increase the quality of online streaming video.

Don't think there won't be content producers - like us - taking advantage of this as well. High definition full-fledged video cameras cost less than $1000 these days. A flash-video "YouTube camera" can record 720p HD for less than $200.

The way to prepare for this is to have good QoS policies in place, so that the day-to-day business data transactions aren't interrupted or slowed when people access online streaming video - which is quickly becoming a necessity.

One big thing that complicates this is hard shut-off date for the end of all analog TV transmissions in the U.S. on February 17, 2009. It is possible to use a converter box to use an older TV with the ATSC standard - but most people will probably get a high definition television instead. High-definition television will prompt high-definition content. That includes home movies.

Right now, High Definition home video cameras are sold to computer geeks, early adopters, and indie filmmakers. They will be more widely adopted when most families have a high definition TV set and want to play back home movies. And as many YouTube videos are harvested from the ranks of home movies, it is possible to then imply that there will be demand for a high definition video hosting service.

That's going to mean more bandwidth usage. QoS policies become crucial.

By Brian Boyko
Editor, Network Performance Daily

We were supposed to have the video of Dr. Steve Fulton we put up on Thursday night up by Wednesday afternoon. That didn't happen.

By coincidence, we also saw this opinion piece by John Dvorak, and linked to it the day before we needed the video to go up. The article, entitled "Don't trust the servers," talks about problems with Windows Genuine Advantage and how it illustrates the problems with SaaS solutions - that you're eternally dependent on a third party to continue to provide service.

In our case, we rely on Google Video to provide the bandwidth and hosting for the videos in our Whiteboard series. Now, you can say what you want about "getting what you pay for" and the like, but when we couldn't upload our video to Google, I realized how dependent I had gotten on their SaaS video hosting solution.

I couldn't log-in to upload the videos. At first I thought the problem was at my end - perhaps Symantec Anti-Virus had caused some sort of conflict or had firewalled off the ports that Google's video uploader needed. That wasn't it. Maybe it was something with our in-house network. That wasn't it either, as I found out when I took a copy of the video home and tried to upload it from both my Windows XP and my Linux partition. Nada. Zip. Zilch.

Despite the fact that Google's help files had a ready - and wrong - answer to every problem I came across, the problem was entirely Google's. And there was nothing I could do about it until they finally fixed whatever the problem was the next day.

See, while there are a number of video hosting solutions out there, Google Video and Veoh were the only ones that allowed me to upload clips more than 10 minutes long - and Veoh's playback was poor. Even if I could re-code the entire video at a lower resolution to lower the filesize, that didn't matter. Ten minutes was the hard limit on YouTube (also owned by Google) and other sites. My 25 minute video needed Google Video.

Eventually I was able to get the video uploaded, and though it took a while to process, it went up last night.

This isn't the first problem we've had with SaaS. Expensable.com often goes down for a few hours and we can't log expense reports. I use Gmail for my personal e-mail and while it's generally reliable, it does have some problems.

All in all, if you're looking at it from a productivity or a network performance view, moving your apps from the local network to a third party service - well, yes, it will absolutely save bandwidth and may make the network run faster for your other apps. But having a faster network doesn't mean anything if the end-user is waiting for a third-party service as long or longer as they used to wait for their slow-loading WAN apps. Or, in other words, you're not solving the problem of slow performance from the perspective of the end-user. You're just shifting blame.

This is not to rag on SaaS. I haven't lost an e-mail since I started using Gmail in 2004. My Flickr account saved pictures of my deceased friend John when my hard drive stopped working one day. And if it wasn't for YouTube, I'd never be able to show my parents in Virginia anything about my life in Texas.

But these experiences are a bit of a wake-up call that SaaS isn't going to solve every problem. Ultimately, the end goal of enterprise network performance is not to keep the network running as fast as possible, or to free up as much bandwidth. It's about finding the solutions which make the end-users more productive and enabling the company to do more as a result.

By Brian Boyko
Editor, Network Performance Daily

Right now there's a bit of a brouhaha about Comcast high speed service. Many Comcast customers are finding themselves cut off from the service because of excessive usage.

To be fair, I was unable to find any reference where Comcast says that their broadband package is "unlimited." However, they fail to disclose what, exactly, "excessive usage" consists of in their Acceptable Use Policy.

I don't have a problem with Comcast limiting bandwidth. There's only so much traffic that their servers can handle, so much that can go down their pipes. Theoretically, limiting the use of the heaviest users would enable better service for the vast majority of users for whom speed is more important than volume.

(Of course, the cynical assume that Comcast is dropping high-usage customers because they're the least profitable and that supporting those users would require investing more in bandwidth and infrastructure - but we'll leave that theory alone for now.)

What I'm concerned about is people suddenly being disconnected from the Internet after passing a line that they know nothing about. I'm sympathetic - my Internet access was cut off without warning back in 1998 at The College of New Jersey, and that cost me a pretty well-paying part-time job as a Web designer. (There were other reasons, but this was a significant reason that I decided to transfer to New Jersey Institute of Technology the next semester.) If my home internet access was cut off today, I'd be at a serious disadvantage with my job editing this blog!

But it also worries me because I can't imagine this happening to a corporate customer. If an IT department asked "how much bandwidth do we have," that information would never be withheld from them. You can't do any meaningful capacity planning if how much capacity you have is kept hidden from you.

Disclosure is obviously the most important step, but there are other options that Comcast could take. Instead of cutting users off, it could throttle down speeds once a customer produces a set amount of traffic - The customer still has access to the Internet, but it doesn't take up quite so much bandwidth. While downloading Linux ISOs via Torrent are going to take longer, viewing YouTube and talking on Skype shouldn't be affected by reasonable, but lower, bandwidth caps.

At any rate, if Comcast simply couldn't keep up with the demand, then perhaps they need to consider billing as a pay-as-you-go service. Sure, we did away with hourly billing around when AOL switched to flat-rate service in 1996… but certainly, paying for the service that you use is probably very appealing to the vast majority of people paying $50 a month to do nothing more than check e-mail and Web browse.

Then again, there are other solutions which are probably preferable. Namely - improving the performance of Comcast's existing infrastructure, or adding capacity to Comcast's existing infrastructure. Apparently, though, both those solutions are more expensive than suddenly dropping a few customers from the rolls and engendering ill-will.

By Brian Boyko
Editor, Network Performance Daily

Recently, World of Warcraft released a patch which enabled players to use integrated VoIP chat. Online gaming and VoIP are, in many ways, extremely well matched. VoIP can help with the immersion of the gaming experience - roleplaying characters with voice, coordinating attacks instantly and in real time, being able to more clearly articulate nuance and inflection that could change the meaning of a sentence… not to mention that you'll finally have some good idea of whether or not the attractive blood elf lady that's been chatting you up is a 45 year old guy living in his mom's basement. Of course, that's not always a good thing…

While I don't play WoW, I have found that VoIP has become a crucial part of my gaming experience - though I typically don't like first-person shooters, I greatly enjoy the Battlefield series because of its interactive voice chat. It's very immersive - with your squad leader barking out orders and relaying info to your commander, it - well, it would be tactless to say that it feels like you're a soldier in war, but it certainly feels like you're a kid playing soldier.

VoIP and gaming are particularly well suited to each other for another reason, more technical and esoteric. VoIP traffic and game traffic usually use the same protocol, UDP.

A quick rundown for the non-technical people reading this post: UDP is a lightweight protocol with no ability to check if a packet was received; TCP is more useful for ensuring that all of the data arrives completely, UDP, that most of the data arrives quickly. This is why UDP is used for online streaming media, voice, and of course, gaming, which requires split-second reflexes and precise timing.

And though we've covered converged data and voice traffic at length before, UDP and TCP on the same network at the same time can cause network and VoIP performance problems if UDP isn't limited to a certain quality of service. Imagine a TCP and UDP connection traveling together. TCP will, in order to make sure that the packets arrive accurately, will slow down its traffic when it senses that there's less room in the pipe. UDP, in order to make sure that the packet arrives quickly, will see that there's now more room in the pipe from what TCP vacated, and take up even more room… which causes TCP to slow even further. It's a vicious cycle.

But voice and data traffic both use UDP - which is one of the reasons that even before WoW's addition of VoIP, people were using Teamspeak or Ventrilo to provide their own voice capabilities with their friends, and though there was almost always a performance hit, the fact that both WoW and Teamspeak are UDP-based makes it easier for both application to co-exist.

There are a few TCP applications in most MMORPG games, but most of them are simple ones - things like transferring inventory and IRC-like chat - which typically don't take up a whole lot of bandwidth compared to the data sent through the game or data sent through the game's VoIP. One thing that IS TCP-based is the downloading of patches and game updates - there are non-technical reasons, such as game balance, that contribute to this, but just about any online game will stop play while you're downloading the patches, rather than downloading the packages in the background while you play. My best guess is that this is partially because coding simultaneous play (UDP) and data download (TCP) is much harder than coding simultaneous play (UDP) and VoIP (UDP.)

One exception to the rule that games must use UDP is Second Life - that MMORPG requires data to be downloaded constantly and accurately, with new items being built. I can't know for sure as I'm not a coder, but I believe this to be one of the reasons why play control (UDP) in Second life tends to suffer so much and objects take a very long time, it seems, to download (TCP).

We'll try to have more technical details on WoW's VoIP rollout later in the week.

This phenomenon isn't limited to the Gummi Bears. There were protests in Mexico over the rising price of corn tortillas. German beers are feeling the pinch as farmers trade in hops and barley for the more lucrative rapeseed and corn. Jolly Time Popcorn isn't feeling so jolly after corn prices went up 70 percent. Between the double whammy of increased cost of corn and increased cost of every other crop because farmers are switching to growing corn instead, it's gotten to the point where it's cheaper to feed livestock, such as pigs, human snack food, according to the Wall Street Journal.

Back to network performance. Consider for the moment, that the path of the food we eat, from raw ingredients to our supermarket shelves consists of a "network" of sorts, this is a classic case of a sudden, unanticipated spike in demand from another endpoint in the network that is wreaking havoc on the network itself. Or, farmers switching to corn at the expense of other crops seems like a classic case of over provisioning one "application" on the network at the expense of all others.

These changes may have seemed insignificant at the time. Many disruptive changes do, which is why you need to have good visibility into your network - whether it's an enterprise network or a food distribution network.

Even a slight increase, for example, in network demand has a number of ancillary costs that many people don't look at. Greater demand of resources doesn't just require more bandwidth. It may necessitate greater processing power, which necessitates more hardware, which necessitates more power and more cooling.

I mean, when even the cost of alternative fuels are going up, many more people are going to be telecommuting from their homes instead of driving or flying in for business. If you're not prepared for a change like that by being aware of how your network is being used, and what changes are coming down the pipe, your network is in just as much trouble as the Gummi Bears are.

By Brian Boyko
Editor, Network Performance Daily

If it sounds like you've read this story before, you probably have. Internet politics in 2000 was newsworthy because, like anyone on AOL, politicians had their own Web pages. In 2004, internet politics were newsworthy because, like anyone with a Blogger.com account, politicians had their own blogs. The new technology of the Web which will "revolutionize politics" this coming election year, but which everyone seems to have found out about two years ago, seems to be YouTube and other "Web 2.0" applications. At least, that's what CNN seems to have focused on with its recent "YouTube" debate of the Democratic Presidential Primaries.

Franken's campaign site will soon include this "Web2.0-style" interactive events map (top) and automatic slideshow (bottom,) seen here in a late beta version. Schlough informed NPD that it would likely be available Friday, August 10, 2007. The events map runs on a combination of Flash and MySQL, while the slideshow runs off of Flash and XML.

Still, that cynicism of overhyped coverage should be reserved not for the politicians but for the news outlets covering the election that grasp onto novelty for novelty's sake. These technologies - Web, blog, and online video - are communications tools. They are important because they very well can help honest politicians communicate with their constituents.

Of course, they can also help dishonest politicians seem to communicate without actually communicating. Most news coverage of the politics of the Internet doesn't actually bother to help people make the distinction.

But what of the people tasked to introduce technology into political campaigns? What is it like to be working on a campaign's IT team in the age when it only recently became clear that a political campaign needs an IT team? We had an opportunity to talk with Jon-David Schlough, who manages IT operations and online communications for the Al Franken campaign for senate in Minnesota.

(Continued...)

Continue reading "The Intersection of Politics and Technology: Talking with Jon-David Schlough, Interactive & IT Operations for the Al Franken for Senate, 2008 campaign." »

Jim McQuaid, senior product manager at NetQoS, talks about how UDP (used for VoIP) and TCP (used for data applications) interact in a short video, as part of our "Whiteboard Series."

If you have questions about the video, please leave a comment below and we'll do our best to answer them.
--------------
More information:

About VoIP:
- VoIP Protocol Basics, and Why VoIP Consumes More Bandwidth Than You Might Expect
- Joe Miller of Linden Labs speaks about VoIP in Second Life
- 'I Swear, There is a Good Reason I'm Singing at Work'

Whitepaper:
- Managing the Performance of Converaged VoIP and Data Applications

Also in our Whiteboard Series:
- The impact of WAN Optimization on TCP Applications
- The impact of WAN Optimization on NetFlow/IPFIX measurements