On 3/29/06, Tom Ziemer <[EMAIL PROTECTED]> wrote: > Hi Leon, > > Spring: Well, our app started as a single jvm version and I used Spring > (and will continue to do so) because it's an incredibly powerful, easy > to use yet unobtrusive framework. You get transaction support, aop, > hibernate/ibatis and support and DI by simply declaring it in your > config file. And the best thing - it actually works - right from the start. > > Apart from that, I don't know anything about Corba and I do NOT think > that my boss would approve if I switched to a technology that none of > our developers has ever used, especially since we're getting close to > the deadline for our project. :o) > > Is RMI performance really that poor? I mean compared to SOAP it still > should be much faster, right? What kind of performance hit are we > actually talking about? Do you have any numbers for Corba/RMI/SOAP ?
I don't have up-to-date comparisons right here, but I would guess that there are always factor 5-10 between the technologies: my guess is: midweight request in corba <1 ms rmi 3-10 ms soap 30 - 100 ms rmi has the disadvantage of java serialization which is slow. SOAP has the big disadvantage of using XML which bloats the request and needs time to remap. Expressed in signal 2 noise ratios I would guess 1:2 for CORBA, 1:10 for RMI and 1:100 for SOAP. > > Regards, > Tom regards Leon --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
