Writers Being Led Astray About CDNs, Wall Street Journal The Latest

Yesterday, the Wall Street Journal published a blog post entitled "Moving Bits through Private Pipes" that talks about content delivery networks and describes how video is delivered. The problem is that like many writers covering the topic, they only get their info from vendors and assume what the vendors are telling them is true. Or they use phrases and words in the post without really knowing how they tie into the technology and don't challenge any of the quotes they get from analysts.

The author says that, "Multimedia communication — particularly video — has caused a strain on public networks, degrading the quality, so privatized networks have become more attractive over the last few years." Where's proof that the public networks are under any "strain"? And what CDNs have "privatized" networks? CDNs do not have "private" networks. They have private cross-connects and things like private peering, but the networks themselves are not "private". They are all delivering content over the Internet, which is public. Yes, some CDNs own the pipes, but the content to the end user is not being delivered from a "private" network.

Berge Avayzian, an analyst with the Yankee Group is quoted in the article as saying, "Video is a huge game changer but its big and bulky, so sending video files over traditional Internet pipes is flawed." What are "traditional internet pipes"? Are there any internet pipes that are un-traditional? And what does Berge classify as being the flaw? The CDN space for video this year alone is north of $400 million, so the "flaw" does not seem to be stopping the market from growing or vendors from growing revenue.

The author of the blog goes on to quote Tata over the recently announced relationship with BitGravity but the statements don't make any sense. Tata is quoted as saying that "BitGravity’s network is designed specifically to handle video," as if they are the only one. Limelight's network was specifically built to handle video from the very first day they launched, years before BitGravity. Tata is also quoted as saying, "The internet as we know is changing, it’s much more of multi-media, communications medium now. Outdated hardware and infrastructure can’t handle that." Who has outdated hardware and infrastructure?

BitGravity and others need to stop coming to the market and saying their content delivery networks are "next-generation" and as a result, Akamai, Limelight and everyone else must be using old, outdated hardware and have a legacy network. What do you think guys like Akamai and Limelight are spending all these CAPEX dollars on? They are not running outdated hardware and their size of their CDN revenue shows that their networks are not outdated. And what about someone like Level 3 who has built out their network in the past 18 months? Does that mean that are using all outdated gear also?

The idea that just because a CDN has been around for more than a few years they are somehow falling behind in being able to deliver content is not accurate. I'd say the opposite. They have more real-world experience building out their network to deliver video and have an advantage over any new CDN who simply does not have the hands-on experience, the volume of traffic or the number of customers that a well-established CDN has.

BitGravity, it's not in your best interest, or the industry's for that matter, to keep trying to portray every other CDN as outdated. If you want the industry to seriously consider your network to be some sort of "next-generation", then put on your website a white paper explaining how it is technically different, with technical details. Because without details, all you are giving us is marketing speak. Looking at the technical page on BitGravity's website gives us no description whatsoever of the technology.

For Bobby White who wrote the article on the WSJ, I feel for him. If CDN is not a topic he tracks and covers on a regular basis, it is very hard for him or any other author to get the straight details, minus marketing speak from many of those in the CDN market. I don't fault someone like Bobby for quality of the blog post he wrote, I fault vendors and analysts in the market who don't take the time to truly explain the business, the technology or the terms in our industry to a writer who is trying to cover our market.

Comments

[Rob]

If I hear about another public network that is "designed to handle video" I'm going to puke. Fibre Channel is designed to handle video, that's about it.

[Vishal]

Can you please explain the rationale behind Tata choosing Bit Gravity? Why wouldn't they take a gamble on Limelight? Cause money is definitely not a problem for the Tata Group.

[Luke]

The BG concept of replicating a video file across all their nodes and storing the full file on each as long as its needed - is an interesting angle for on-demand delivery differing from the traditional caching approach supported by most CDNs. This is potentially compelling for speedy delivery of long tail (video) content, however looks like it creates as many issues as it solves in terms of the additional storage it requires. Some clarity on how they solve this catch 22 could be a great selling point and differentiator?! Tata can certainly help them out with HSIP, pipes and data centres - perhaps the $11.5m investment from Tata will be sunk straight back into buying network services from Tata?!

[Steve Lerner]

"replicating a video file across all their nodes and storing the full file on each as long as its needed - is an interesting angle"

It is a very old angle.

I used the same technique to build the TEN-TV media CDN patent#6,801,576 (now owned by Nokia) and the Speedera CDN media services (now owned by Akamai).

Limelight then built their CDN this way, and many others have since followed.

Everyone has been doing it this way for the last eight years. I am continually amazed at the amnesia in the media and the industry... None of the features under discussion have any correlation's with CDN's success.

There is only one common thread among CDNs that are/were profitable and survived: performance.

-Steve

[Ken]

And BitGravity has always performed

[Roger Williams]

Well done Dan, the FUD is strong in this industry. While it matures we are going to have to deal with more traditional writers trying to get their head around this stuff. Keep up the fight for clarity.

The concept of keeping all content in cache all the time is not exclusive to BitGravity. What is exclusive to Bit Gravity is that they offer it as the default setup. My issue with this is: Why should you pay to have content in cache when its not being requested?

The other CDN's can apply this service as well but it will cost you extra as it requires more resources.

[Martin Hayward]

Core CDN vendors who’ve been in the business of delivering digital media for years have the experience in contructing their own networks as well as the ability to deliver content.

[PK]

Well I have to agree completely. I think everyone, well almost everyone, in our industry wants to see clear, factual writing from any and all media coverage. While it can be said that hype and spin are present in any sector of any industry, this article is spot on. Perhaps the hype around companies who get funding and have yet to deliver packet one can also be put into perspective.