DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://nagoya.apache.org/bugzilla/show_bug.cgi?id=11369>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=11369

Poor performance under JDK 1.4.1_01





------- Additional Comments From [EMAIL PROTECTED]  2002-08-01 16:15 -------

I've made a 20-second test run under OptimizeIt on Solaris with JDK 1.4.0_01
64-bit, and this is the strange things I see:

In a 20 second run, one particular thread shows:

Only 3.66 seconds (18%) waiting for the server to respond in
java.net.SocketInputStream.read(). This should generally be the most important
time consumer, and it's not.

5.59 seconds (28%) in java.net.URLEncoder.encode() called from
UrlConfig.getQueryString(). This is a ridiculous amount of time for such an
operation: obviously part of the problem.

2.12 seconds (in JMeterThread.delay(). Taking into account that I don't have any
delays in my script, it looks like too much.

0.67 seconds (11%) in java.lang.StringCoding.encode(), called from
HTTPSampler.getResponseHeaders(). This may be legitimate, though.

The same test under the 1.3.1_02 JDK shows:

26% waiting for the server to respond. Still not much, but definitely better.
24% in JMeterThread.delay(). WHY?
Only 4.6% in URLEncoder.encode().

So looks like the difference lies mostly in the URLEncoder.encode(). The fun
thing is that I don't need (nor want) my parameters URL-encoded!

--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

Reply via email to