My understanding is that it is the creation of an Exception that is expensive (because of the calculation of stacktrace information ?). The actual throwing is less expensive. Isn't it that log4j uses the stacktrace data to intelligently mine information. So even though it never throws the exception the overhead of creation is still incurred.
..then again I could be wrong :) Tim. ----- Original Message ----- From: <[EMAIL PROTECTED]> To: "Log4J Users List" <[EMAIL PROTECTED]> Sent: Monday, February 02, 2004 10:10 AM Subject: RE: performance problems IBM JVM Ceki G�lc� <[EMAIL PROTECTED]> wrote on 02/02/2004 12:26:08 PM: > > Log4j computes location information by throwing an exception which is > caught immediately and parsing the stack trace of the exception. > Apparently, creating an exception on AIX is slower than on other platforms. This is a bit of a nitpick, but it may prevent others from needlessly looking throught the code. The terms used in your description worried me for a moment, in terms of efficiency. To clarify: It seems, from looking at source, you don't actually throw an exception, and hence there is nothing to catch. Rather you do the more sensible thing which is to simply instantiate an exception. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
