Re: less invocation time with larger data size
John McClain <[email protected]>
| Newsgroups | gmane.comp.java.sun.jini |
|---|---|
| Message-ID | <[email protected]> |
fatma elshehri wrote: > The average time of the call in one client case reached 1350 ms while > for two clients case for the same data size it has reached 850 ms (in > some cases), for the same data size. 1350 ms, that is to say 1.3 seconds? Depends on your hardware and network of course, but these days RMI "null calls" (no work done in the server, no args, and a void return) are generally less than millisecond. Do you know how many bytes are in the return value? The other side of this is how much time is each call spending in the server? > Entries of the returned array are of the same class, which is not so > complicated it has ,mainly two dimentional array to keep records > selected and one other array to keep soem metadata and two simple > functions. There are duplications in the array elements inside When an object appears more than once in a serialization stream 2nd and subsequent appearances are replace by a 5 byte (I think its 5 bytes - something on that order) "pointer" to the first appearance. This can result in quite a bit fewer bytes being sent. If the 500 record case was really 5 copies of the 100 record case it might end up being reprsented by a similar number of bytes because of this. Hope this helps -------------------------------------------------------------------------- Getting Started: http://www.jini.org/wiki/Category:Getting_Started Community Web Site: http://jini.org jini-users Archive: http://archives.java.sun.com/archives/jini-users.html Unsubscribing: email "signoff JINI-USERS" to [email protected]