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.

By Brian Boyko
Editor, Network Performance Daily

My roommates and I have been chosen (if we choose to accept it,) to be a Nielsen family. And while, at first, I was thrilled to finally get the opportunity to stand up and be counted (I'm a fan of more cancelled series than I can count - Firefly, Brisco County Jr., Wonderfalls, The Critic, Futurama, Studio 60 on the Sunset Strip, Arrested Development, Farscape, The Tick, Justice League Unlimited…) it got me thinking about how, exactly, Nielsen is going to get accurate data on their version of "network performance."

First of all, my "Nielsen Family" consists of three roommates, including myself, who share a four bed, four-bath condo. I've never seen Patrick watch television. Ever. Mike supposedly watches TV but I've never seen him actually pay attention to it - preferring to leave it humming in the background to help him concentrate while he plays Warcraft III or Diablo II.

As for me, well, that's the stickiest of wickets. I own a TV but I only use it for my PS2 and Wii hookups. While Nielsen can track what people watch on TiVo brand PVR boxes, I'm not sure they'll be able to track my setup, which consists of a P4 Dell repurposed into a MythTV on Ubuntu box. The MythTV box connects directly to a business-class projector, but I end up using it mostly for DVD movies, YouTube, and downloaded videos. Even when I watch MythTV recorded movies, I usually end up doing so from my bedroom, which can access the MythTV box and stream the videos (and live TV!) over the LAN in the condo.

(Continued...)

Continue reading "Another kind of "Network monitoring,": Looking at the Nielsen Ratings." »

The big problem with WAN Optimization and Application Performance Monitoring was that there simply wasn't a WAN Optimization solution on the market that preserved end-to-end performance data, nor a monitoring solution that would work in an optimized WAN.

This problem has been solved.

At Cisco Networkers in Anaheim, NetQoS gave a presentation to hundreds of attendees to make the announcement that we've been working with Cisco to develop a management interface for accurate end-to-end application response time measurement that works on optimized networks. (In addition to the people mobbing our booth, John Chambers, CEO of Cisco Systems, stopped by for a chat and review of what we do.)

Through integrated software on Cisco Wide Area Application Services devices (WAAS), TCP header information is exported to NetQoS SuperAgent (an end-to-end application performance monitoring module) before optimization occurs - preserving that information. Finally, IT organizations can accurately validate the results of WAN Optimization deployments.

(Continued…)

Continue reading "Tracking The Optimized WAN: NetQoS Integrates with Cisco WAAS to Deliver End-to-End Application Response Time Reporting for WAN Optimization" »

Julie Bort interviewed George Kurian at Cisco in Network World, where they talk about WAN optimization.

The interview talks about how Cisco's optimization and acceleration products are distinguished from competitors' and (of course, considering that George Kurian works for Cisco) promoted as superior because of their transparent placement in the network. This means they can be shared among several servers and applications, as well as integrated with Cisco's existing products, and QoS and security policies do not have to be migrated or disrupted. One item only barely touched upon is the idea of using a single appliance in the branch office - the Integrated Service Router - to handle WAN optimization, security, and routing - and how having one appliance to handle all these tasks helps cut down on server room clutter and complexity.

To be sure, these appear to be advantages to Cisco's solution. But the dirty little secret is that all WAN optimization solutions on the market, including Cisco's, obscure end-to-end performance metrics. This is a major issue, of course, and makes the current state of choosing whether or not to deploy a WAN optimization solution a Monty Hall problem - do you opt to retain visibility into your network performance and the ability to solve problems faster, or you deploy a WAN optimization device and hope that whatever's behind curtain number two (the resulting performance gain) is better than what you've traded for?

Maintaining transparency of response time and latency metrics is critical in our view and any WAN optimization vendor that provides a solution to this problem will have a serious competitive advantage.

By Ben Erwin, Product Manager for NetQoS Performance Center 3.0
& Steve Harriman, Vice President of Marketing at NetQoS:

We've recently held a Webinar to discuss NetQoS Performance Center 3.0. We had a Q&A session there and we would like to reproduce some of the questions attendees asked about NetQoS Performance Center 3.0, and the answers we provided. We also plan to answer the questions we couldn't get to during the Webinar on this blog by the end of this week.

(Continued...)

Continue reading "Q&A on NetQoS Performance Center 3.0" »

By Brian Boyko

Back in November of 2006, (which seems like such a long time ago,) Network Performance Daily published a column by Carol Schiraldi about "why enterprise developers use Java and game programmers use C++."

We published this for a number of reasons - but the main one was that typically, enterprise developers are programming for function first, reliability, second, and performance over the network, if it's even considered at all, is a tertiary thought.

What this means is that applications, developed originally for the LAN environment, often take up valuable network resources unnecessarily when placed into a WAN environment.

[Full disclosure: NetQoS sells network performance management software which diagnoses problems like "chatty apps," and we want you to buy them. Anyway…]

But one area where this isn't a significant problem is in game development, which was the thesis of the original column. Game developers, who realize their games have to perform well over the Internet, typically build with performance in mind first.

This was confirmed when we had a chance to talk to Timothee Besset, a game developer at ID Software, developers of the famous Wolfenstein, Doom, and Quake series of games. Here's what he said about this issue:

(Continued…)

Continue reading "ID Software Developer Timothee Besset on Network Performance in Games" »

By Brian Boyko

If you spend some time poring through RFC documents (something I don’t recommend for the 99% of the population that is still sane) you’ll find tons of improvements, modifications, case-specific optimizations and alternatives to TCP, the workhorse of networking transport protocols since the 1970s.

Seth Noble, President and Founder of Data Expedition, Inc., believes that he can do one better. His company claims that their proprietary transport protocol, MTP, for “Multipurpose Transaction Protocol,” provides a scalable alternative to TCP that uses bandwidth more efficiently. According to Mr. Noble:

“TCP's 1970's data model makes dealing with this problem more difficult than it needs to be. TCP was created with the assumption that packet loss would "rarely" occur, and so it is rather fragile in our modern, congested networks. A lot of very smart people have tried for many years to patch TCP and help it cope, but it still carries its 30 year old legacy with it.”

“MTP/IP was designed from scratch to operate in congested environments where packet loss and other network problems are common. As a result, MTP/IP does an exceptionally good job of quickly AND correctly identifying whether or not data has really been lost and then recovering that data with little or no disruption.”

(Continued...)

Continue reading "Proprietary MTP: an alternative to TCP?" »

By Brian Boyko

"Jesus has come back, and he's a phone now."

The Apple iPhone might not be the second coming, but it will certainly be popular. Already the Web is being inundated by speculation, information, mis-information, rumor, innuendo, and anyone with an opinion on Apple writing about the iPhone.

So we thought: Why buck the trend?

The truth is that when this device comes out, many people are going to buy the iPhone, they will use it at their jobs - including those in a corporate IT environment - and that means it is going to become the responsibility of the IT manager.

(Continued...)

Continue reading "Why Apple's iPhone means more work for the IT department" »

by Jeff Hicks

Voice over IP (VoIP) is a hot topic in enterprise networking - mostly because it's a challenge. In implementation, VoIP employs a number of different protocols, and has a unique set of performance requirements that make it a challenge for any data network. Examining VoIP protocols should give someone a basic idea of the performance requirements that VoIP places on the network.

First, there's call setup, which sets up everything needed to make the telephone connection between the caller and the recipient (or “callee”). This requires protocols that enable dial tone, number lookup, ringing, and busy signals before the call even occurs. In addition, the call setup protocols handle things that happen after the call - any resource cleanup and statistic reporting.

Call setup protocols use either TCP or UDP to transfer data during the setup and takedown phases of a telephone call. The messages are sent back and forth between the caller, recipient, and call server using well-known ports. For calls that travel between the VoIP network and the Public Switched Telephone Network (PSTN), the call server will converse with a VoIP gateway using the call setup protocol. There are many different call setup protocols, some standardized and some proprietary. Let’s discuss a few of these.

(Continued...)

Continue reading "VoIP Protocol Basics, and Why VoIP Consumes More Bandwidth Than You Might Expect." »

Prof. Christopher Yoo, Vanderbilt University School of Law

Professor Christopher Yoo joined the faculty of the Vanderbilt University School of Law in 1999, and his research focuses primarily on how technological innovation and economic theories of imperfect competition are transforming the regulation of electronic communications.

In addition to clerking for Justice Anthony M. Kennedy and working at the law firm of Hogan & Hartson under the supervision of now-Chief Justice John G. Roberts, Jr., he has also published "Network Neutrality and the Economics of Congestion" [PDF] in Georgetown Law Journal, and "Beyond Network Neutrality" [PDF] in the Harvard Journal of Law and Technology.

Thanks to Network Performance Daily for giving me the chance to discuss some of my ideas and to the various readers on Network Performance Daily and Slashdot.com for their comments. The comments suggesting that QoS and optimizing networks have nothing to do with network neutrality are a testament to how much the network neutrality debate has changed over the last year and a half. For example, the network neutrality Policy Statement issued by the FCC in August 2005, which remains the focus of the network neutrality conditions in the AT&T-BellSouth, Verizon-MCI, and SBC-AT&T mergers, it focuses on providing differential service on the basis of application and devices as well as content source.

(Continued...)

Continue reading "Net Neutrality: Professor Yoo Responds to Comments" »