Application Delivery Market

Maybe I should get a tattoo

Dan Matte — Thu, 10 May 2007 23:59:05 GMT

Gartner recently published a new Magic Quadrant for Application Delivery Products (which basically means BIG-IP for F5).

If I ever decide to get a tattoo, the new magic quadrant has to be on the short list of designs. We're the only vendor to have always been in the leader quadrant since Gartner started publishing a Magic Quadrant for this category in 2001. The whole F5 team should be very proud of this achievement. Of course, since Gartner's criteria keep the quadrant drifting up and to the right over time, we can't sit on our laurels if we want to be in a similar position next go around.

The new Magic Quadrant for Application Delivery Products can be found here.

Tax Relief from Spam

Dan Matte — Fri, 16 Mar 2007 07:00:43 GMT

There have been some interesting email threads flying around internally over the past week discussing spam and how people must deal with it. You likely know that we formed a partnership with Secure Computing that resulted in a product offering called the Message Security Module (MSM) that can be licensed on a BIG-IP. The announcement is located here.

The email threads were related to MSM and discussed how long mail had to be retained and what constituted actually receiving a message which, in turn, leads to it having to be retained.

Companies are increasingly facing requirements to store email for longer periods of time. There is a huge amount of resources wasted on storing and backing up spam messages that have absolutely nothing to do with business. Once you've received the message, you need to deal with it - even if it is spam and that costs both time and money. Intel is currently facing a challenge to come up with old emails and I'm sure system administrators everywhere hate getting the request "we need to find all email associated with _____". If spam is eliminated from what had to be retained in the first place, at least there would be less to sift through.

We’re hearing of organizations even backing up the spam that their spam gateways process. The end users don’t see it hit their inbox but the internal systems still have to deal with it and that just doesn't seem right.

Typical anti-spam devices must receive, parse, scan and then evaluate a message in order for them to determine what to do with it (nicely put, Ken). Blocking spam by never allowing the message to be received in the first place seems to be a very good way to negate all taxes that downstream systems have placed on them due to spam - lower load on the network, less load on the anti-spam gateways, less storage required, less bandwidth for data backup, etc. With MSM, email from known bad senders is blocked before the message is ever received.

MSM basically works this way:

Send mail (SMTP) request is made from the Internet to BIG-IP running MSM
We do a query to Secure Computing's TrustedSource reputation database
TrustedSource replies with a reputation score for the sender
We either terminate the SMTP connection before the sender is able to send the message or pass it through based on the reputation score and the thresholds configured.
Done

Since the spam message is never actually received, other systems don't have to deal with it and thereby enjoy "Tax Relief" from the burden that spam would otherwise place on them.

F5 #1 According to Gartner

Dan Matte — Tue, 12 Dec 2006 15:15:00 GMT

Gartner recently published their quarterly market share measurements for the Application Delivery Controller market. We're very fortunate to again occupy the number one spot. This is pretty amazing especially given the caliber of companies (Cisco, Juniper, Citrix, Foundry and more) that we're up against.

Thanks to all our customers, partners and F5ers who help make this happen every day!

Citrix CFO provides some insight to Netscaler business

Dan Matte — Mon, 11 Dec 2006 15:45:00 GMT

Some interesting tidbits from a few recent financial conferences where David Henshall, Citrix's CFO, presented:

At the Credit Suisse First Boston annual Technology Conference November 30, 2006:

BV Jagadeesh (former CEO of Netscaler) is moving to "more of an evangelical role" from his GM role for the Application Networking Group
The Western US Sales region has been responsible for about half of their Netscaler revenue
They are moving the Western regional sales VP into Jagadeesh's former GM role

At a Goldman Sachs Software and IT Services Conference on November 8, 2006:

WANScaler products (Orbital acquisition) will be launched in January 2007
Addressing channel partner adoption of WANScaler he said "It should be a lot easier to adopt than the Netscaler product."

What's to be made of this? Here's my take:

In 6 months it will have been 2 years since Citrix acquired Netscaler. It looks like BV Jagadeesh is positioning himself to move out of the company. Not surprising and likely coincides with financial handcuffs coming off.
Concentration of sales in the Western US - this is consistent with what we've seen in the market. Peel out a couple of key accounts and there are very few sales in other areas. In their last earnings call they also said that there was essentially no Netscaler revenue for APAC and a little from EMEA. It appears that broad adoption has been a challenge for them.
Ever since making the Netscaler acquisiton, Citrix has been touting the leverage they would be able to achieve with their traditional channel partners. From our observations, we believe that the majority of Netscaler revenue continues to come from the original Netscaler resellers, not the traditional Citrix channel partners. Mr. Henshall's comment seems to confirm this. I also think they will discover that the WANScaler adoption won't be "a lot easier".

Citrix's market share in the Application Delivery Controller space continues to hover between 7-8%. I don't believe that this latest round of changes will materially improve their prospects to gain market share.

Truthiness in Vendor Claims

Dan Matte — Tue, 17 Oct 2006 13:43:00 GMT

You may have heard of the term "truthiness" coined by Stephen Colbert, host of the fake news show "The Colbert Report".

I read a blog entry on Network Computing's site that uses that term in reference to vendor performance claims. At the heart of the discussion that ensues with the vendor is the inability to reproduce the performance claims that the vendor makes in a real environment.

At F5 we take that very seriously. Over time we have been accused of being too conservative with how we characterize the performance of our devices. Some of the vendors that we compete against make their performance claims based on truthiness. A couple of examples are using stats like TCP connection setups per second or what we term infinite to infinite connections.

The TCP setup per second stat ignores the fact that there is also work to be done to tear the connection down let alone actually send data across the connection - both of which dramatically reduce the original number claimed. Typically vendors that have some type of hardware acceleration that is good at setting up connections will cite that stat. Several years ago, Alteon Websystems, now owned by Nortel, used that form of measurement extensively to make big performance claims. More recenly, we've seen the TCP setup per second claim resurface in the launch materials surrounding Cisco's ACE. Kudos to Cisco for bringing back a classic.

Citrix's Netscaler product performance claims use what we call infinite to infinite connections. Basically that is multiplexing as many individual requests as possible over relatively few connections between the client and the Netscaler box as well as between the Netscaler box and the server. That's something you can do in the lab but that you would never actually see in the field. We're seeing more customers who were initially lured by the high performance claims switching to F5 when reality did not match the expectations that were set.

We don't use those tricks when we come up with our performance claims I would rather provide data that is useful to customers to make an informed sizing decision. Will we always win at every possible performance metric that can be cooked up in a lab? No. But I believe that customers appreciate truth over truthiness and that will serve the market better in the long term.

Some False Claims by Citrix

Dan Matte — Wed, 20 Sep 2006 14:32:00 GMT

Citrix announced a new version of software for their Netscaler product yesterday. They made a claim in their press release that is simply not true. They stated: "This latest release of Citrix NetScaler extends our leadership in this market and further advances state-of-the-art web application delivery."

There are two things about that statement that need to be corrected.

First, they are claiming to extend their leadership in the market. Hmmmm...do they lead in market share? No. Do they lead in functionality? No. Do they lead in performance? No. Are they rated highest by industry analysts? No. Leader? I think not. Extending a lead? Absolutely not!

Second, the feature set that they are introducing in this new version is a subset of what we introduced to the market when we released version 9 two years ago! So, they're not advancing the state of the art either - escpecially since we've added functionality and extended performance to BIG-IP since then.

Press releases are used to tout new things but we shouldn't allow companies to make claims that are simply false without subjecting them to scrutiny.