[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 Philippe Mouawad changed: What|Removed |Added Resolution|WONTFIX |DUPLICATE --- Comment #19 from Philippe Mouawad --- *** This bug has been marked as a duplicate of bug 63055 *** -- You are receiving this mail because: You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 Philippe Mouawad changed: What|Removed |Added Status|NEEDINFO|RESOLVED Resolution|--- |WONTFIX Keywords||FixedInTrunk --- Comment #18 from Philippe Mouawad --- pmoua...@apache.org 8:45 AM (1 minute ago) to commits Author: pmouawad Date: Tue Dec 4 07:45:20 2018 New Revision: 1848126 URL: http://svn.apache.org/viewvc?rev=1848126=rev Log: Bug 62887 - Customized SampleLabel is ignored and overwritten by JMeter 5.0 Document incompatible changes of JMeter 5.0 Bugzilla Id: 62887 Modified: jmeter/trunk/xdocs/changes_history.xml -- You are receiving this mail because: You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 --- Comment #17 from Philippe Mouawad --- (In reply to Alan Silva from comment #16) > (In reply to Philippe Mouawad from comment #10) > > (In reply to Alan Silva from comment #9) > > > Oh yes. You're absolutely right @Philippe Mouawad. > > > > > > The patch should haven been an if/else. My mistake. > > > > > > In any case, as I said in my previous comment: > > > > > > Calling addSubResult(SampleResult subResult, false) is good enough I > > > think. > > > > > > That's what I ended up doing and I would presume that it's what other > > > Jmeter > > > users that upgraded to 5.0 and use custom result labels did as well. > > > > Hello Alan, > > Thanks for your feedback. > > > > So In your opinion, is it ok to do nothing ? or should we do something. > > > > Of course we should add a note on this in release notes. > > > > Thanks > > Hi, sorry I stopped monitoring this ticket for a bit. > > IMO, you could leave the current implementation as is, as long as it's > mentioned in the rel notes like you suggested. > > As long as other users like myself don't have to open up the code to find > out why things changed, then I think everything is good. > > Cheers Thanks Alan for your feedback. @Artem, thanks a lot for your work on this and PR. But, IMO, we shouldn't do anything about it except add to release notes of 5.0 the Incompatible changes section. -- You are receiving this mail because: You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 --- Comment #16 from Alan Silva --- (In reply to Philippe Mouawad from comment #10) > (In reply to Alan Silva from comment #9) > > Oh yes. You're absolutely right @Philippe Mouawad. > > > > The patch should haven been an if/else. My mistake. > > > > In any case, as I said in my previous comment: > > > > Calling addSubResult(SampleResult subResult, false) is good enough I think. > > > > That's what I ended up doing and I would presume that it's what other Jmeter > > users that upgraded to 5.0 and use custom result labels did as well. > > Hello Alan, > Thanks for your feedback. > > So In your opinion, is it ok to do nothing ? or should we do something. > > Of course we should add a note on this in release notes. > > Thanks Hi, sorry I stopped monitoring this ticket for a bit. IMO, you could leave the current implementation as is, as long as it's mentioned in the rel notes like you suggested. As long as other users like myself don't have to open up the code to find out why things changed, then I think everything is good. Cheers -- You are receiving this mail because: You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 --- Comment #15 from Artem Fedorov --- Benchmark results: # Run complete. Total time: 00:37:59 Benchmark (index) Mode Cnt Score Error Units StringBenchmark.compareToBenchmark0 thrpt 20 0.823 ± 0.013 ops/ns StringBenchmark.compareToBenchmark1 thrpt 20 0.829 ± 0.007 ops/ns StringBenchmark.compareToBenchmark2 thrpt 20 0.822 ± 0.012 ops/ns StringBenchmark.compareToBenchmark3 thrpt 20 0.822 ± 0.017 ops/ns StringBenchmark.compareToBenchmark4 thrpt 20 0.817 ± 0.012 ops/ns StringBenchmark.compareToBenchmark5 thrpt 20 0.829 ± 0.003 ops/ns StringBenchmark.compareToBenchmark6 thrpt 20 0.828 ± 0.009 ops/ns StringBenchmark.equalsBenchmark 0 thrpt 20 0.829 ± 0.002 ops/ns StringBenchmark.equalsBenchmark 1 thrpt 20 0.745 ± 0.006 ops/ns StringBenchmark.equalsBenchmark 2 thrpt 20 0.745 ± 0.006 ops/ns StringBenchmark.equalsBenchmark 3 thrpt 20 0.818 ± 0.013 ops/ns StringBenchmark.equalsBenchmark 4 thrpt 20 0.738 ± 0.013 ops/ns StringBenchmark.equalsBenchmark 5 thrpt 20 0.744 ± 0.008 ops/ns StringBenchmark.equalsBenchmark 6 thrpt 20 0.828 ± 0.003 ops/ns StringBenchmark.compareToBenchmark0 avgt 20 1.208 ± 0.009 ns/op StringBenchmark.compareToBenchmark1 avgt 20 1.218 ± 0.017 ns/op StringBenchmark.compareToBenchmark2 avgt 20 1.236 ± 0.021 ns/op StringBenchmark.compareToBenchmark3 avgt 20 1.203 ± 0.005 ns/op StringBenchmark.compareToBenchmark4 avgt 20 1.211 ± 0.011 ns/op StringBenchmark.compareToBenchmark5 avgt 20 1.234 ± 0.029 ns/op StringBenchmark.compareToBenchmark6 avgt 20 1.208 ± 0.012 ns/op StringBenchmark.equalsBenchmark 0 avgt 20 1.203 ± 0.003 ns/op StringBenchmark.equalsBenchmark 1 avgt 20 1.353 ± 0.024 ns/op StringBenchmark.equalsBenchmark 2 avgt 20 1.342 ± 0.008 ns/op StringBenchmark.equalsBenchmark 3 avgt 20 1.209 ± 0.003 ns/op StringBenchmark.equalsBenchmark 4 avgt 20 1.341 ± 0.015 ns/op StringBenchmark.equalsBenchmark 5 avgt 20 1.351 ± 0.019 ns/op StringBenchmark.equalsBenchmark 6 avgt 20 1.209 ± 0.014 ns/op # Run complete. Total time: 00:37:54 Benchmark (index) Mode Cnt Score Error Units StringBenchmark.compareToBenchmark0 thrpt 20 0.802 ± 0.028 ops/ns StringBenchmark.compareToBenchmark1 thrpt 20 0.807 ± 0.032 ops/ns StringBenchmark.compareToBenchmark2 thrpt 20 0.802 ± 0.022 ops/ns StringBenchmark.compareToBenchmark3 thrpt 20 0.794 ± 0.043 ops/ns StringBenchmark.compareToBenchmark4 thrpt 20 0.808 ± 0.019 ops/ns StringBenchmark.compareToBenchmark5 thrpt 20 0.826 ± 0.007 ops/ns StringBenchmark.compareToBenchmark6 thrpt 20 0.807 ± 0.011 ops/ns StringBenchmark.equalsBenchmark 0 thrpt 20 0.822 ± 0.011 ops/ns StringBenchmark.equalsBenchmark 1 thrpt 20 0.743 ± 0.007 ops/ns StringBenchmark.equalsBenchmark 2 thrpt 20 0.722 ± 0.026 ops/ns StringBenchmark.equalsBenchmark 3 thrpt 20 0.815 ± 0.016 ops/ns StringBenchmark.equalsBenchmark 4 thrpt 20 0.736 ± 0.019 ops/ns StringBenchmark.equalsBenchmark 5 thrpt 20 0.719 ± 0.023 ops/ns StringBenchmark.equalsBenchmark 6 thrpt 20 0.815 ± 0.017 ops/ns StringBenchmark.compareToBenchmark0 avgt 20 1.221 ± 0.013 ns/op StringBenchmark.compareToBenchmark1 avgt 20 1.258 ± 0.058 ns/op StringBenchmark.compareToBenchmark2 avgt 20 1.250 ± 0.044 ns/op StringBenchmark.compareToBenchmark3 avgt 20 1.254 ± 0.061 ns/op StringBenchmark.compareToBenchmark4 avgt 20 1.229 ± 0.022 ns/op StringBenchmark.compareToBenchmark5 avgt 20 1.216 ± 0.016 ns/op StringBenchmark.compareToBenchmark6 avgt 20 1.273 ± 0.075 ns/op StringBenchmark.equalsBenchmark 0 avgt 20 1.244 ± 0.049 ns/op StringBenchmark.equalsBenchmark 1 avgt 20 1.376 ± 0.045 ns/op StringBenchmark.equalsBenchmark 2 avgt 20 1.377 ± 0.040 ns/op StringBenchmark.equalsBenchmark 3 avgt 20 1.220 ± 0.032 ns/op StringBenchmark.equalsBenchmark 4 avgt 20 1.348 ± 0.010 ns/op StringBenchmark.equalsBenchmark 5 avgt 20 1.375 ± 0.041 ns/op StringBenchmark.equalsBenchmark 6 avgt 20 1.232 ± 0.040 ns/op So 'equals()' looks a little bit slower than 'compareTo()', but the difference between them is not bigger than ~10% -- You are receiving this mail because: You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 --- Comment #14 from Artem Fedorov --- Created attachment 36260 --> https://bz.apache.org/bugzilla/attachment.cgi?id=36260=edit Benchmark test class I created a benchmark test for equals and compareTo methods. Dependencies: org.openjdk.jmh jmh-core 1.19 org.openjdk.jmh jmh-generator-annprocess 1.19 -- You are receiving this mail because: You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 --- Comment #13 from Artem Fedorov --- (In reply to Felix Schumacher from comment #12) > (In reply to Artem Fedorov from comment #11) > > Hi > > I added a PR https://github.com/apache/jmeter/pull/429 > > Also I use compareTo() instead of equals() for string comparison. > > Performance comparison of these functions I got from > > http://jniosocket.sourceforge.net/result.html > > Have you tried to reproduce those performance experiments? They were done > with java 1.4.2 which is reallly old. Sadly the source code used for those > experiments is not available anymore (at least not from the site directly). Sure, this code and experiment is very old and not actual for current Java version. The String implementation was changed (I found 1.4 jdk and decompile code). I can try to make benchmark test and attach some results late -- You are receiving this mail because: You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 --- Comment #12 from Felix Schumacher --- (In reply to Artem Fedorov from comment #11) > Hi > I added a PR https://github.com/apache/jmeter/pull/429 > Also I use compareTo() instead of equals() for string comparison. > Performance comparison of these functions I got from > http://jniosocket.sourceforge.net/result.html Have you tried to reproduce those performance experiments? They were done with java 1.4.2 which is reallly old. Sadly the source code used for those experiments is not available anymore (at least not from the site directly). -- You are receiving this mail because: You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 --- Comment #11 from Artem Fedorov --- Hi I added a PR https://github.com/apache/jmeter/pull/429 Also I use compareTo() instead of equals() for string comparison. Performance comparison of these functions I got from http://jniosocket.sourceforge.net/result.html -- You are receiving this mail because: You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 --- Comment #10 from Philippe Mouawad --- (In reply to Alan Silva from comment #9) > Oh yes. You're absolutely right @Philippe Mouawad. > > The patch should haven been an if/else. My mistake. > > In any case, as I said in my previous comment: > > Calling addSubResult(SampleResult subResult, false) is good enough I think. > > That's what I ended up doing and I would presume that it's what other Jmeter > users that upgraded to 5.0 and use custom result labels did as well. Hello Alan, Thanks for your feedback. So In your opinion, is it ok to do nothing ? or should we do something. Of course we should add a note on this in release notes. Thanks -- You are receiving this mail because: You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 --- Comment #9 from Alan Silva --- Oh yes. You're absolutely right @Philippe Mouawad. The patch should haven been an if/else. My mistake. In any case, as I said in my previous comment: Calling addSubResult(SampleResult subResult, false) is good enough I think. That's what I ended up doing and I would presume that it's what other Jmeter users that upgraded to 5.0 and use custom result labels did as well. -- You are receiving this mail because: You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887
--- Comment #8 from Philippe Mouawad ---
(In reply to Artem Fedorov from comment #7)
> Hi all,
>
> I agree with attached fix, but it can throw NPE, because it skip check for
> null
>
>
> if(subResult == null) {
> // see https://bz.apache.org/bugzilla/show_bug.cgi?id=54778
> return;
> }
>
> The other suggestion is add jmeter.property for enable/disable subResults
> rename option.
Hello Artem,
1/ IMO, having a property that can break report is a bit weird no ? And it's
another option for a requirement that is a edgy.
2/ Regarding the patch , I don't understand why addSubResult is called twice.
Am I missing something ? Should the second addSubResults be in an else ?
Same issue for other methods unless I am wrong.
3/ Also wouldn't this patch have an impact on JMeter performances by adding a
String comparison for every Sub result addition .
Thanks
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887
--- Comment #7 from Artem Fedorov ---
Hi all,
I agree with attached fix, but it can throw NPE, because it skip check for null
if(subResult == null) {
// see https://bz.apache.org/bugzilla/show_bug.cgi?id=54778
return;
}
The other suggestion is add jmeter.property for enable/disable subResults
rename option.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 62887] Customized SampleLabel is ignored and overwritten by JMeter 5.0
https://bz.apache.org/bugzilla/show_bug.cgi?id=62887 Philippe Mouawad changed: What|Removed |Added OS|Linux |All Summary|SampleLabel displays|Customized SampleLabel is |incorrectly in Jmeter 5.0 |ignored and overwritten by ||JMeter 5.0 Hardware|PC |All -- You are receiving this mail because: You are the assignee for the bug.
