On Sun, Sep 14, 2008 at 4:47 PM, Kent Närling <[EMAIL PROTECTED]> wrote: >> Considering this is (presumably) over a network connection, then this >> latency looks to me to be quite good. But I wonder how this was > > > No, actually it was not! > I simply initialized a (mockup) webservice , published it and then called it > from within the same process > (tried both with single thread and multiple threads with comparable results)
The web service I talk about bellow, has two responsibilities: authenticating user cridentialts and crud-like user administration. Naturally, you need to be authenticated to administrate users, so the admin part of the service knows about the authentication service through the same Java interface that CXF uses (JAXB or somethin' front-end) and then through Spring I can inject the actual service implementation object into the admin. service. This way, it dosn't have to go through CXF or any XML marshalling in order authenticate its users - just ordinary calls to a normal object through an interface. But I don't know if that is applicable for your situation - I suppose it isn't since you started this discussion. > >> Only a proper load-test can tell whether or not the throughput will be >> good enough. My experience is that you throughput-bottleneck is much >> more likely to be in your application code that in CXF itself. At >> least, that is my conclusion for load that ranges up to 20.000 >> requests per minute. > > > This is a very interesting comparison!! 333 TPS? > Can you qualify that a bit more? is that on a single machine? I had to do a little digging in my e-mail archive to find any confirmed numbers, but I did find some. We did a load test on the 28'th of December, 2007 on an SSO web service we were building (using newest CXF at the time). The server was a virtual dual-core linux 2.6 machine running a JBoss on a Sun Java 5 JVM. This web service was doing user authentication and communicated with two load-balanced LDAP servers. Using JMeter, we had 90 clients pound it with a total of 6.000 requests, and got these numbers: + average response time: 87 ms. + median: 59 ms. + 90% line: 125 ms. + Min. response time: 11 ms. + Max response time: 4.423 ms. + Throughput: 287.2 Req/sec + KB/sec: 171.4 -- Venlig hilsen / Kind regards, Christian Vest Hansen.
