NetScaler Scores in Content Switch Test

Light Reading today published the results of the first independent test of Layer 7 content switches, devices that promise to help carriers boost their profitability by offering premium grade Internet services to corporate customers, at the same time as cutting data center costs (see LR Completes Content Switch Test).

The tests demonstrate that the performance of this type of equipment has come on by leaps and bounds in the past year or so. Full details are in the report on Light Testing: Content Switch Test.

Three vendors -- Extreme Networks Inc. (Nasdaq: EXTR), NetScaler Inc., and WinCom Systems Inc. -- submitted products for the tests, which were conducted by the European Advanced Networking Test Center AG (EANTC) using the Ixia 1600T performance analyzer, running the IxWeb content switch test suite from Ixia (Nasdaq: XXIA). In the tests, all three vendors' products handled massive volumes of simultaneous HTTP sessions and TCP connections. They also demonstrated high resilience to distributed denial of service (DDOS) attacks and packed a big punch on secure sockets layer (SSL) performance.

"While the first 'Web switches' on the market a few years ago were basically enhanced Layer 4 (TCP port) switches, these systems are highly specialized, focusing on all aspects of the complex science of HTTP switching," writes Carsten Rossenhövel, EANTC's managing director, research and manufacturer testing, in the test report. All three switches in the test "are ready for the busiest (and most attacked…) Web servers, with enough reserves for the foreseeable future," Rossenhövel adds.

This is no mean feat, bearing in mind that content switches have to identify applications in individual sessions and then switch them. Their ability to do this on a massive scale, handling hundreds of thousands of HTTP sessions at the same time, lies behind hopes that they'll help carriers roll out lucrative new secure Internet services to business users. It also lies behind hopes that content switches will enable carriers to build bigger, more efficient data centers.

In the Light Reading test, content switches were subjected to three groups of performance tests: HTTP performance (session rate and capacity, latency and load balancing), TCP performance (connection rate and capacity), and security (SSL performance and resilience to DDOS attacks). In addition, the price and feature set of each switch was compared.

A marking system was used that enabled Light Reading and EANTC to demonstrate the way in which overall scores were calculated for a typical content switch application.

NetScaler got the highest overall score: 4.3 out of a possible 5. It got the best average mark for each group of tests, although it didn't come out on top in every individual test. Its most remarkable results were handling close to 2.4 million simultaneous HTTP sessions or TCP connections -- and having a list price that's half that of Extreme's and WinCom's.

Extreme and WinCom tied for second place, each getting an overall score of 3.5 out of 5. Extreme did well, but not exceptionally well, in all tests. WinCom had the best switching rate (both HTTP and TCP), but didn't do so well on latency and capacity tests. This is largely because its switch is based on a different architectural concept, one that aims to use the Internet itself as a TCP connection buffer.

The tests were commissioned and paid for entirely by Light Reading, ensuring the integrity of the results (see Testing Testing).

— Peter Heywood, Founding Editor, Light Reading
Page 1 / 3   >   >>
Hanover_Fist 12/5/2012 | 12:43:48 AM
re: NetScaler Scores in Content Switch Test Where or where is little Foundry Networks in this test?
hey_tedd 12/5/2012 | 12:43:47 AM
re: NetScaler Scores in Content Switch Test This story sounds like a sales add!
Iipoed 12/5/2012 | 12:43:47 AM
re: NetScaler Scores in Content Switch Test Why bother, they have thousands of chassis and stackable L4-7 switches installed over the last 5 years. They are the marketshare leaded in 4-7 ports shipped/installed and have one of the most if not the most mature product in this market. What possible benefit would they derive from this test.

Where was F5, Alteon/Nortel, Radware, Cisco?
Hanover_Fist 12/5/2012 | 12:43:46 AM
re: NetScaler Scores in Content Switch Test Blah...blah...blah...

Isn't this interesting - when the shoe is on the other foot, Floundry instantaneously points out "where was vendor x in this test," but when push comes to shove, they end up shoveling the same sh*t line of "why bother...yada yada yada..."
new_light 12/5/2012 | 12:43:46 AM
re: NetScaler Scores in Content Switch Test They are the marketshare leaded in 4-7 ports shipped/installed and have one of the most if not the most mature product in this market.
I love it when people try to put a positive spin on something that is obviously negative. When you say "mature", aren't you really saying their product is old out dated...
Iipoed 12/5/2012 | 12:43:45 AM
re: NetScaler Scores in Content Switch Test Why do you really care whether they were in the test or not. There are certainly bigger issues for this site to deal with than ex employees bitterness and negative comments which really are of no interest to 99% of the people that post here. Get a life and add something worthwhile to this group.
Hanover_Fist 12/5/2012 | 12:43:44 AM
re: NetScaler Scores in Content Switch Test Hey Mr. Ex-Foundry Sales Guy,
If you believe in Floundry sooooo much, why did you bolt the day you reached your 4-year vesting date (as you stated on a previous post)? Guess you loved Bobby so much, you felt you could do much better retired than selling, hmmmm?

I guess life wasn't all apple pie and roses on that side of the fence?
Iipoed 12/5/2012 | 12:43:44 AM
re: NetScaler Scores in Content Switch Test See my previous response. Talk to the customers, they are not going to buy netscaler, or the other two. Why bother. Who in their right mind is going to buy from a startup or a vendor that has tried three different manufacturers to align themself with and still not gain ground. Extreme-tried to push f5, alteon, some other company than no longer is in business and now some proprietary technology.

You obviously only know how to put out negative comments with no foundation at all to back them up. Get a life.
digerato 12/5/2012 | 12:43:43 AM
re: NetScaler Scores in Content Switch Test There are only two considerations if you are a vendor considering entering a product into a bake-off. The second question is only relevant if you answer yes to the first:

1) Do you need the exposure?
2) Will you win given the test criteria?

If you're an established vendor with an established product range, you don't need to enter a bake-off. You have customer validation and success stories, so why risk showing yourself up?

If you need the exposure, either because you're a new or small vendor, or you have a brand new box that customers have not accepted yet, then you take a close look at the testing criteria and consider if you are willing to put the effort behind ensuring your box wins. If you're confident your box is going to do well on the test criteria, then you enter.

This second consideration is often given the least thought. Or, if the vendor is really very desperate for any exposure at all, there's lots of finger-crossing and hoping you come out on top.

If the answer to question (2) is "No", this is when you consider paying Kevin Tolly to do a test for you (where you're guaranteed to win).

And now, back to the "my box is better than your box" squabble.


Iipoed 12/5/2012 | 12:43:43 AM
re: NetScaler Scores in Content Switch Test I had enough respect for Foundry and their future growth to step aside and let new blood with greater motivation have a wonderful opportunitly.

Also 7 days in the hospital, additional health issues and what the hell the ability to retire at 50 without a care in the world other than to goof with you characters that post silly stuff on this site.

Hey ask yourself this about BJ. He has million and millions yet continues to focus on building a better company and better products. He does not need to but he does out of committment to his employees, customers and the industry. It is not an ego trip either. Having sat in many meetings with him I never saw an egotisically person merely an engineer that believed in what he was pushing for Foundry to become.
Page 1 / 3   >   >>
Sign In