You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@jmeter.apache.org by bruto <bo...@gmail.com> on 2014/03/01 12:23:30 UTC
sample times reported by JMeter seem to scale incorrectly with the
number of threads
Hello everyone
tried searching this mailing list for a bit, couldn't find a relevant answer
Using JMeter 2.11 (20140227) to load test our Mirth Connect installation
(HL7 frontend)
Got a simple test plan with a thread group that sends a single request to
Mirth via a TCP sampler, 10 times in a loop
Using both 'view results in table' standard listener and 'response times
over time' plugin from jmeter-plugins to check the results
Max connections on Mirth side is set to way more than 1 or 3, HL7 message
ids are generated randomly for each sample, so there're no collisions there
(and they're sent using different sockets anyway)
1 thread:
JMeter side: table <http://imgbox.com/iFc089tn> and chart
<http://imgbox.com/CbD2GoPH>
Mirth side: http://imgbox.com/1cgAgi0g <http://imgbox.com/1cgAgi0g>
response times line up pretty well so far
now the same test plan, but with 3 threads, ramp-up period 3s:
table <http://imgbox.com/AtsgkBny> chart <http://imgbox.com/i16OCBNf>
mirth <http://imgbox.com/PzD0t4c1>
It's as though JMeter sums up response times from all threads, what's up
with that?
Thanks in advance to whoever might shed light on this
--
View this message in context: http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455.html
Sent from the JMeter - User mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
For additional commands, e-mail: user-help@jmeter.apache.org
Re: sample times reported by JMeter seem to scale incorrectly with
the number of threads
Posted by sebb <se...@gmail.com>.
On 1 March 2014 12:24, bruto <bo...@gmail.com> wrote:
> It can cope with the load just fine, this is merely to check that the test
> plan stats work
>
> anyway, tried running 3 JM instances, each with 10 reps in 1 thread,
> starting at the same instant, on schedule
> the first one shows this: http://imgbox.com/pypFZMym
> <http://imgbox.com/pypFZMym>
> the other two show read timeouts
> very weird
> tried checking and unchecking 're-use connection' in all three - same result
>
I have tried running multipe TCP samplers in parallel against a
webserver and do not see the problems you are experiencing.
>
> --
> View this message in context: http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455p5719457.html
> Sent from the JMeter - User mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
> For additional commands, e-mail: user-help@jmeter.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
For additional commands, e-mail: user-help@jmeter.apache.org
Re: sample times reported by JMeter seem to scale incorrectly with
the number of threads
Posted by bruto <bo...@gmail.com>.
It can cope with the load just fine, this is merely to check that the test
plan stats work
anyway, tried running 3 JM instances, each with 10 reps in 1 thread,
starting at the same instant, on schedule
the first one shows this: http://imgbox.com/pypFZMym
<http://imgbox.com/pypFZMym>
the other two show read timeouts
very weird
tried checking and unchecking 're-use connection' in all three - same result
--
View this message in context: http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455p5719457.html
Sent from the JMeter - User mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
For additional commands, e-mail: user-help@jmeter.apache.org
Re: sample times reported by JMeter seem to scale incorrectly with
the number of threads
Posted by sebb <se...@gmail.com>.
On 1 March 2014 11:23, bruto <bo...@gmail.com> wrote:
> Hello everyone
>
> tried searching this mailing list for a bit, couldn't find a relevant answer
> Using JMeter 2.11 (20140227) to load test our Mirth Connect installation
> (HL7 frontend)
>
> Got a simple test plan with a thread group that sends a single request to
> Mirth via a TCP sampler, 10 times in a loop
> Using both 'view results in table' standard listener and 'response times
> over time' plugin from jmeter-plugins to check the results
> Max connections on Mirth side is set to way more than 1 or 3, HL7 message
> ids are generated randomly for each sample, so there're no collisions there
> (and they're sent using different sockets anyway)
>
> 1 thread:
> JMeter side: table <http://imgbox.com/iFc089tn> and chart
> <http://imgbox.com/CbD2GoPH>
> Mirth side: http://imgbox.com/1cgAgi0g <http://imgbox.com/1cgAgi0g>
>
> response times line up pretty well so far
>
> now the same test plan, but with 3 threads, ramp-up period 3s:
> table <http://imgbox.com/AtsgkBny> chart <http://imgbox.com/i16OCBNf>
> mirth <http://imgbox.com/PzD0t4c1>
>
> It's as though JMeter sums up response times from all threads, what's up
> with that?
Most likely the server cannot cope with the additional load.
But I suppose it's possible that there's some kind of JMeter bug - or
a bug in the script - that is causing the samples to interfere with
each other.
A simple way to test if this is the case is to use 3 copies of JMeter
each running a single thread and see what happens.
> Thanks in advance to whoever might shed light on this
>
>
>
> --
> View this message in context: http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455.html
> Sent from the JMeter - User mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
> For additional commands, e-mail: user-help@jmeter.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
For additional commands, e-mail: user-help@jmeter.apache.org
Re: sample times reported by JMeter seem to scale incorrectly with
the number of threads
Posted by Philippe Mouawad <ph...@gmail.com>.
Thank you
On Mon, Mar 3, 2014 at 8:11 PM, Philippe Mouawad <philippe.mouawad@gmail.com
> wrote:
> Hello,
> Could you then close the bug as invalid as it wast obvious to me, I
> understood that the echo server used for bug report was single threaded,
> but that original "issue"was still
> valid.
>
> Thank you
>
> On Monday, March 3, 2014, bruto <bo...@gmail.com> wrote:
>
>> I already pointed out my own mistake in the first comment to that issue :)
>> Thanks for taking the time to look at it
>>
>>
>>
>> --
>> View this message in context:
>> http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455p5719475.html
>> Sent from the JMeter - User mailing list archive at Nabble.com.
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
>> For additional commands, e-mail: user-help@jmeter.apache.org
>>
>>
>
> --
> Cordialement.
> Philippe Mouawad.
>
>
>
>
--
Cordialement.
Philippe Mouawad.
Re: sample times reported by JMeter seem to scale incorrectly with
the number of threads
Posted by Philippe Mouawad <ph...@gmail.com>.
Hello,
Could you then close the bug as invalid as it wast obvious to me, I
understood that the echo server used for bug report was single threaded,
but that original "issue"was still
valid.
Thank you
On Monday, March 3, 2014, bruto <bo...@gmail.com> wrote:
> I already pointed out my own mistake in the first comment to that issue :)
> Thanks for taking the time to look at it
>
>
>
> --
> View this message in context:
> http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455p5719475.html
> Sent from the JMeter - User mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org <javascript:;>
> For additional commands, e-mail: user-help@jmeter.apache.org<javascript:;>
>
>
--
Cordialement.
Philippe Mouawad.
Re: sample times reported by JMeter seem to scale incorrectly with
the number of threads
Posted by bruto <bo...@gmail.com>.
I already pointed out my own mistake in the first comment to that issue :)
Thanks for taking the time to look at it
--
View this message in context: http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455p5719475.html
Sent from the JMeter - User mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
For additional commands, e-mail: user-help@jmeter.apache.org
Re: sample times reported by JMeter seem to scale incorrectly with
the number of threads
Posted by sebb <se...@gmail.com>.
On 3 March 2014 10:23, bruto <bo...@gmail.com> wrote:
> Added some logging to the vanilla TCPSampler (using vanilla TCPClientImpl),
> it appears that protocolHandler.read() execution time is to blame
> It's a different socket, different SocketInputStream every time, but still
> some synchronization must be going on causing the response read time to
> inflate
I don't see how it can be a problem with the TCP sampler.
The implementation is extremely simple; if the read takes a long time
it must be a problem outside the JMeter code.
Possibly in the JVM, or network, or server.
As I already wrote, try running multiple separate JMeter instances and
see if the times increase.
Ideally on separate hosts - but beware of using VMs on the same
physical host as these can interfere with each other.
>
>
> --
> View this message in context: http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455p5719468.html
> Sent from the JMeter - User mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
> For additional commands, e-mail: user-help@jmeter.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
For additional commands, e-mail: user-help@jmeter.apache.org
Re: sample times reported by JMeter seem to scale incorrectly with
the number of threads
Posted by bruto <bo...@gmail.com>.
Added some logging to the vanilla TCPSampler (using vanilla TCPClientImpl),
it appears that protocolHandler.read() execution time is to blame
It's a different socket, different SocketInputStream every time, but still
some synchronization must be going on causing the response read time to
inflate
--
View this message in context: http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455p5719468.html
Sent from the JMeter - User mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
For additional commands, e-mail: user-help@jmeter.apache.org
Re: sample times reported by JMeter seem to scale incorrectly with
the number of threads
Posted by bruto <bo...@gmail.com>.
there: https://issues.apache.org/bugzilla/show_bug.cgi?id=56206
--
View this message in context: http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455p5719467.html
Sent from the JMeter - User mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
For additional commands, e-mail: user-help@jmeter.apache.org
Re: sample times reported by JMeter seem to scale incorrectly with
the number of threads
Posted by sebb <se...@gmail.com>.
On 2 March 2014 12:29, bruto <bo...@gmail.com> wrote:
> this is not the kind of load to cause the server response times to increase
> I posted Mirth screenshots for a reason, they show consistent response times
> from server's point of view in both scenarios
> I still think it's some sort of calculation error on Jmeter's side, just
> need to find out where it's coming from
If you can provide a simple test case that the JMeter developers can
use, then please create a Buzilla issue and attach it with details of
the results you are seeing.
Please note that it should only use standard JMeter test elements.
>
>
> --
> View this message in context: http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455p5719461.html
> Sent from the JMeter - User mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
> For additional commands, e-mail: user-help@jmeter.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
For additional commands, e-mail: user-help@jmeter.apache.org
Re: sample times reported by JMeter seem to scale incorrectly with
the number of threads
Posted by bruto <bo...@gmail.com>.
this is not the kind of load to cause the server response times to increase
I posted Mirth screenshots for a reason, they show consistent response times
from server's point of view in both scenarios
I still think it's some sort of calculation error on Jmeter's side, just
need to find out where it's coming from
--
View this message in context: http://jmeter.512774.n5.nabble.com/sample-times-reported-by-JMeter-seem-to-scale-incorrectly-with-the-number-of-threads-tp5719455p5719461.html
Sent from the JMeter - User mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
For additional commands, e-mail: user-help@jmeter.apache.org