Yan, Axis2 indeed translates REST request into SOAP. However, this doesn't mean that REST is a priori slower since in contrast to a SOAP request, there is no XML parsing involved in this translation.
Actually, for both protocols, Axis2 parses the request into a SOAP infoset. Since a REST request has a simpler structure, one would expect that it is faster, but probably the difference is so small that other factors than parsing might be dominant. I think there is really no a priori argument to predict what will be faster and you will have to do a performance test to find out. Andreas On Sun, Aug 9, 2009 at 22:04, Yan Liu<[email protected]> wrote: > Hi, > > I'm not clear how Axis2 implemented REST. But if I evaluate response time of > a REST call and a SOAP call to the same Axis2 web service deployed on the > same machine (with REST enabled), which one will be faster? REST is supposed > to be lighter and faster. Am I right? The main reason I ask is that I was > wondering whether the REST implementation in Axis2 is just translating http > get request into a soap message and passing it to the SOAP engine for > processing. If that is the case, REST call would be slower. > > I thought I should ask this before actually doing the performance test to > find it out. > > Thanks, > Yan > > > >
