This is the mail archive of the java@gcc.gnu.org mailing list for the Java project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

RMI Netwoek Overhead


Hi!

We are making a master thesis and in that we are comparing overhead sizes on 
method calls
using RMI,RMI-IIOP, CORBA, DCOM and .Net Remoting. In our tests we have found 
that RMI generates the second largest amount of network traffic after .NET 
Remoting with SOAP on a single method call with a byte as parameter.
It uses almost  three times the amount as .NET Remoting in binary mode. Can 
someone explain why it is like this.

API 	             Byte 	Int 	Long 	Double
------------------------------------------------------
RMI 	   	      852    855  859 	        859
RMI-IIOP 	437 	432 	444 	444
Remoting TCP 	299 	301 	306 	308
Remoting HTTP 	1722 	1717 	1722 	1732
CORBA 		497 	500 	504 	504
DCOM TCP 		
DCOM UDP 	

We have found that when sending primitive data type arrays as parameters the 
network traffic is not
linear. Is there some optimization on primitive data type arrays in different 
sizes?

Thanks in advance!  Im not on the mailing list so please reply to my mail 
adress.

Best regards
   Per Arneng

Attachment: mail.txt
Description: Text document


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]