Hmmm... In all fairness I'm guessing DataPower has something similar so if I hear about or discover anything of this sort for their product I will post it.
In the mean time this report seems quite interesting and worthy of a read for sure.
[Update: In further review of this the following seems like an important detail to take notice to. Very Interesting Stuff!]
---
"Using real world test cases, we tested XML Web services security and
acceleration at very high transaction rates and widely diverse sizes using a
single Sarvega appliance against ten dual-processor Pentium 4 Xeon servers.
The results of the Sarvega products demonstrated new levels for performance
not before seen from a single product." said Kevin Cooke, Director of
Independent Testing Services."
---
Hey Sarvega: Did I mention I would be more than happy to run independent tests of my own and publish them here on XSLTBlog? You don't have to thank me... it would be my pleasure ;) :D
TrackBack URL for this entry:
http://www.xsltblog.com/xslt-blog-mt/mt-tb.cgi/176
You do realize how these studies are done, right? Company X figures out the absolute best scenario to showcase their particular piece of software/hardware and, optionally, the very worst for their competitors. They carefully design a benchmark to show their product in the best possible light. Then they pay an “independent” research company to run the exact same tests on the exact same configurations they’ve already run, so they know in advance what the results will be. Even then, sometimes they’ve missed a crucial variable, and the results don’t come out as expected, in which case they either don’t publish the results, or figure out what went wrong and take a second swing at the pinata with more carefully specified rules.
It’s completely unscientific and pointless. It proves nothing except that benchmarks can be tailored to demonstrate anything you want to demonstrate.
Elliotte,
This is exactly the type of comment I was hoping to draw from someone with the right to demand the respect of those who choose to listen. My hope when I post things of this nature is to draw in people to comment and as such bring out more of whats really going on behind the scenes.
Obviously someone who actually owns the hardware or has a better feel for what reality generally tends to be in cases of this nature (yourself for example) can follow up the post with a “uh, those aren’t the results we’re getting…” or “yeah, we bought one and we like it but…” or “blah, blah, blah… same bullsh— different marketing department.” Doing so will help bring the truth, whatever that may be, to those who are paying attention.
XML optimization is obviously going to be a hot topic sometime soon. It may die just as fast as it started if through experience and conversation its determined that theres no interest or no need. If this interest exists and the need for solutions becomes obvious hardware acceleration is something I think people will easily be drawn to, especially when reports of this nature hit the airwaves. Discovering whats really going on behind the scenes is something we obviously need to determine. I don’t think any of us in the professional XML development world want people to suddenly turn sour to XML because of one product or another not solving the problems it claims to solve. Hopefully your comments will help avoid any lashback if it turns out this is just another piece of hype from another “clever” marketing department.
Thanks for bringing this out Elliotte! Would you mind if I brought attention to your comment via a quick post and a link or would you prefer to let people either find it themselves or lead them here via a posting/trackback from your blog? Either way, your comments are appreciated :)
Cheers!
<M:D/>