You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@jmeter.apache.org by bu...@apache.org on 2013/07/29 11:55:32 UTC
[Bug 55325] New: Negative value in average column of Aggregate
Report
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
Bug ID: 55325
Summary: Negative value in average column of Aggregate Report
Product: JMeter
Version: 2.8
Hardware: PC
OS: Windows XP
Status: NEW
Severity: normal
Priority: P2
Component: Main
Assignee: issues@jmeter.apache.org
Reporter: uddirulz@gmail.com
Created attachment 30645
--> https://issues.apache.org/bugzilla/attachment.cgi?id=30645&action=edit
Jtl file
Hi,
I am using WebService(SOAP) requests to load test a web service for 400 users
and duration of 5 minutes.On starting the test,after some time the average
column in the aggregate report starts showing negative value.The response time
increases and turns negative.
I have attached the jtl file.
Please provide a solution on how to get rid of it.
Thanks,
Uddi
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
Philippe Mouawad <p....@ubik-ingenierie.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |NEEDINFO
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #13 from Sebb <se...@apache.org> ---
Have you tried setting the JMeter property
sampleresult.useNanoTime=false
on the host where the test runs?
Can you provide a copy of the jmeter.log file? From both server and client
please.
You may need to redact it for sensitive information (server names etc.)
Can you provide the JMX file you use?
Again, redact any private settings first.
Also, try running the test locally, not via client-server mode.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #4 from uddirulz@gmail.com ---
I have upgraded to Jmeter 2.9.And removed the additional plugins. But the
matter still persists.
After some time negative values are getting generated in the aggregate report.
Thanks
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
uddirulz@gmail.com changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEEDINFO |NEW
--- Comment #19 from uddirulz@gmail.com ---
I have attached the jmx file and the jtl file without any 3rd party plugins.
-ve results are still persisting.
And i have searched for sampleresult.useNanoTime=false in the properties file
but couldnt find any such thing. :(
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #11 from Sebb <se...@apache.org> ---
Yes, there may be issues with nanotime on specific systems.
But if it were a general problem I would expect a lot more complaints.
When using nanotime, JMeter regularly (default every 5s) recalculates its
offset from systemTime, so even if the nanotimer is static it should not behave
as we are seeing.
Also, the start time is calaculated using the same clock type, and the start
time appears to be OK in the result file - it is incrementing normally.
It's looking much more like a buggy 3rd party sampler.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
Philippe Mouawad <p....@ubik-ingenierie.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEEDINFO |RESOLVED
Resolution|--- |WORKSFORME
--- Comment #24 from Philippe Mouawad <p....@ubik-ingenierie.com> ---
Closing as user did not provide required information.
Cannot do anything without it.
Feel free to reopen with required information.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #15 from Philippe Mouawad <p....@ubik-ingenierie.com> ---
*** Bug 55337 has been marked as a duplicate of this bug. ***
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
uddirulz@gmail.com changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEEDINFO |NEW
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #23 from Sebb <se...@apache.org> ---
(In reply to uddirulz from comment #22)
> True that the server is getting pent up with requests.I have tried with
> fewer threads.Its working fine.But my requirement is to test with large no.
> of threads.
In that case, maybe try with two JMeter instances, each with half the load.
Or three, etc.
Perhaps try disabling the samplers that don't actually get run.
Also try removing the unnecessary Transaction Controllers.
> Other samplers present in the jmx file r not getting executed in the
> specified duration of the test run so they are not present in the jtl file.
> My question is even if the server is overwhelmed with requests why will
> negative response time come?
That's the question. We don't know what the problem is. But we still don't have
all the information we have requested. We still need the JMeter log file that
corresponds with the JTL file you have provided.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #5 from Philippe Mouawad <p....@ubik-ingenierie.com> ---
Hello,
Can you attach jmeter.log file ?
Are you using distributed testing ?
Thanks
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #8 from Rainer Jung <ra...@kippdata.de> ---
I haven't looked at the code, but could it be that the clock on the system was
set back? For instance on some virtualizations the clock can jump quite a bit
during normal operations, because simulating CPU interrupts is too expensive
and NTP only adjusts small clock jitter. So if the clock gets out of sync by
too much an external apparatus often lets the clock jump to the correct time.
Is this system virtualized? If not, does it use NTP successfully?
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #2 from uddirulz@gmail.com ---
I have added jmeter plugin 1.1.0.
Latest version is also giving the same issue.Can you provide a solution?
Thanks
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #21 from Sebb <se...@apache.org> ---
The JTL file still does not appear to correspond with the JMX file.
The JMX file contains more samplers than there are samples in the JTL file.
Also, we need the jmeter.log file that corresponds with the test run.
The Test Plan itself has lots of unusual features (though these should not
cause the issue):
- a loop count of 1 with a thread count of 400 won't produce useful test
results
- the Transaction Controllers only have a single child sampler, so they don't
do anything useful.
- the BeanShell pre-processor concatenates strings multiple times.
I suggest you try running the test with far fewer threads. It looks like the
server is being overwhelmed with requests, otherwise why are the responses
taking longer and longer?
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #7 from Sebb <se...@apache.org> ---
Sorry, forgot to say: the negative value in the Aggregate Report is only the
symptom of the problem.
The actual problem is that some elapsed times are negative.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
uddirulz@gmail.com changed:
What |Removed |Added
----------------------------------------------------------------------------
Severity|normal |major
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #9 from Sebb <se...@apache.org> ---
I don't think a changed clock could affect more than a few samples, because the
time is refetched per sample. Unless all the negative samples started before
the clock was reset, which seems unlikely as the negative times started at
04:28:17 and lasted until the end at 04:31:31.
But there is another oddity in the elapsed times - they seem to get bigger and
bigger, starting at 40308 and reaching 755182 just before the negative times
start. That's a ratio of about 15:1 which is very strange.
It might just possibly be worth trying disabling the nano-second time
resolution by setting the JMeter property:
sampleresult.useNanoTime=false
I suppose if the nano-second clock is very badly behaved it might cause some
issues, but I'm not sure how the negative times can come about.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #12 from uddirulz@gmail.com ---
I am connecting to a remote desktop and running the test there. I have not seen
the clock getting affected as such.Running the test even with the simple thread
group is giving the same negative values.
Dont know what is the issue!! :(
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #17 from uddirulz@gmail.com ---
Created attachment 30672
--> https://issues.apache.org/bugzilla/attachment.cgi?id=30672&action=edit
The jmx file of the test
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #10 from Rainer Jung <ra...@kippdata.de> ---
(In reply to Sebb from comment #9)
> It might just possibly be worth trying disabling the nano-second time
> resolution by setting the JMeter property:
>
> sampleresult.useNanoTime=false
>
> I suppose if the nano-second clock is very badly behaved it might cause some
> issues, but I'm not sure how the negative times can come about.
There are some indications that the nano second timers can be problematic:
http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6458294
http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6440250
and an article about how it's supposed to work:
https://blogs.oracle.com/dholmes/entry/inside_the_hotspot_vm_clocks
Some of this indicates that on Windows XP before SP2 a CPU local counter is
used that is not synchronized between cores and can vary in frequency between
cores due to power management. So when threads are switched between cores bad
things can happen. This specific problem should have been fixed in Windows XP
SP 2.
Measuring time is really something complicated :(
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #16 from Philippe Mouawad <p....@ubik-ingenierie.com> ---
Can you provide your Test Plan and also generated JTL without Third party
plugin ?
This will help investigation.
Also did you try setting this in user.properties:
sampleresult.useNanoTime=false
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #14 from Philippe Mouawad <p....@ubik-ingenierie.com> ---
*** Bug 55337 has been marked as a duplicate of this bug. ***
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #6 from Sebb <se...@apache.org> ---
The attached JTL file shows that the elapsed times start out OK, but after a
certain point (Line 1058) all of them become negative.
It also shows that the test seems to be using the 3rd party Stepping Thread
Group.
Whether or not that is the cause is not clear at present, but it's certainly a
possibility.
If you can provide a simple test case that uses standard JMeter code only, that
will be a lot easier to investigate.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
Philippe Mouawad <p....@ubik-ingenierie.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |NEEDINFO
--- Comment #3 from Philippe Mouawad <p....@ubik-ingenierie.com> ---
Thanks but that's not what was required which is:
- upgrade to jmeter 2.9
- remove jmeter-plugins
Thanks
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #18 from uddirulz@gmail.com ---
Created attachment 30673
--> https://issues.apache.org/bugzilla/attachment.cgi?id=30673&action=edit
The jtl file without ant 3rd party plugins
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
Philippe Mouawad <p....@ubik-ingenierie.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |NEEDINFO
CC| |p.mouawad@ubik-ingenierie.c
| |om
--- Comment #1 from Philippe Mouawad <p....@ubik-ingenierie.com> ---
It seems you are using a third party plugin.
Could you try without it and with last version 2.9 of JMeter ?
Thank you
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
uddirulz@gmail.com changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #30672|0 |1
is obsolete| |
--- Comment #20 from uddirulz@gmail.com ---
Created attachment 30674
--> https://issues.apache.org/bugzilla/attachment.cgi?id=30674&action=edit
The jmx file of the test
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 55325] Negative value in average column of Aggregate Report
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55325
--- Comment #22 from uddirulz@gmail.com ---
True that the server is getting pent up with requests.I have tried with fewer
threads.Its working fine.But my requirement is to test with large no. of
threads.
Other samplers present in the jmx file r not getting executed in the specified
duration of the test run so they are not present in the jtl file.
My question is even if the server is overwhelmed with requests why will
negative response time come?
--
You are receiving this mail because:
You are the assignee for the bug.