Hi Philippe,
I can confirm that RC3 fixes my problem and I now have stable memory usage
having run the same .jmx overnight for confirmation.
Thanks for your assistance with this matter. Case closed! :)
--
View this message in context:
Thanks for the prompt response.
- No other test elements in use other than what I listed in the OP.
- Jmeter version: 2.5
I've tried removing all the listeners and running from NON-GUI mode with -l
from the master, this is the command I used: ./jmeter -n -r -t cli-test.jmx
-l example-log
Sadly
Hi Philippe, thanks for your guidance.
I can confirm I was using the HTTPClient 4 implementation. Would you expect
the Java implementation to avoid this problem?
I've downloaded the latest binary from the nightly releases but am
struggling to get jmeter-server to run, it bombs on start with the
I have the http://code.google.com/p/jmeter-plugins/ plugin pack in place the
master server which I was using for some graphing listeners but I removed
them during testing to ensure they were not the burden.
I have no additional jars in place on the slave servers.
I'll retest with HTTP client v3
Early results look like the HttpClient 3 is increasing its memory usage at a
much slower rate than the HttpClient 4 jmx was doing.
I am still using dist however, did you intend for me to test HttpClient 3 on
2.5.1-RC3?
Additionally, I seem to be getting an increased error rate with HttpClient
3,
Hi, I'm trying to run a fairly simple test plan in distributed mode and
experiencing some interesting memory consumption issues which I hope someone
can shed some light on..
The problem is regardless of how many threads I use, each of the slave
servers sees growing memory usage until they
6 matches
Mail list logo