You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@jmeter.apache.org by Stefan Adamovsky <sp...@gmx.de> on 2006/11/28 10:35:28 UTC
90% Time higher much higher than the Average???
Hi
I am wondering that I have a 90% Time with 4,1 seconds and a Average time with 2,5 seconds. How is this possible? Normally the 90% line has to be under the response time of the average! (90% time means that the 10% of samples with the highest response time are ignored, or am I wrong?)
There are a few samples (Max-time) with extrem high response with about 1 800 000 seconds (~30min) and I have 0.05%Errors. Is it possible that the average is calculated with the errors and 90% calculated without? Or can anybody eplain me why this happens?
regards
Stefan
--
Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!
Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: 90% Time higher much higher than the Average???
Posted by Peter Lin <wo...@gmail.com>.
usually the 90% line is higher than the average time.
the 90% line is the line where 90% of the response time finished.
peter
On 11/28/06, Stefan Adamovsky <sp...@gmx.de> wrote:
>
> Hi
> I am wondering that I have a 90% Time with 4,1 seconds and a Average time
> with 2,5 seconds. How is this possible? Normally the 90% line has to be
> under the response time of the average! (90% time means that the 10% of
> samples with the highest response time are ignored, or am I wrong?)
> There are a few samples (Max-time) with extrem high response with about 1
> 800 000 seconds (~30min) and I have 0.05%Errors. Is it possible that the
> average is calculated with the errors and 90% calculated without? Or can
> anybody eplain me why this happens?
>
> regards
> Stefan
> --
> Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!
> Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>
>
RE: 90% Time higher much higher than the Average???
Posted by "Christensen, Alan" <al...@hp.com>.
The best way to think about both median and 90% time is to imagine that the test has finished and that the samples have been placed in order based upon response time. The median would be computed by counting halfway through the list and taking the response time of the sample at that location. For fairly uniform response time distributions, the median will be close to the average. The 90% time would be found by counting 90% of the way through the list (from fastest to slowest) and then taking the response time of the sample at that location. Because it is computed this way, 90% of the samples will have response times less than or equal to than the 90% time. Hence the 90% time will usually be significantly higher than the average or median times.
The aggregate listener needs to keep around all the measured values for all the samples (and sort them on the fly) in order to compute the 90% and median values. If you don't need median and 90% values, then you are much better off to use the summary report, as it requires a lot less memory and processing when the test runs.
-----Original Message-----
From: Stefan Adamovsky [mailto:spam.adam@gmx.de]
Sent: Tuesday, November 28, 2006 1:35 AM
To: jmeter-user@jakarta.apache.org
Subject: 90% Time higher much higher than the Average???
Importance: High
Hi
I am wondering that I have a 90% Time with 4,1 seconds and a Average time with 2,5 seconds. How is this possible? Normally the 90% line has to be under the response time of the average! (90% time means that the 10% of samples with the highest response time are ignored, or am I wrong?) There are a few samples (Max-time) with extrem high response with about 1 800 000 seconds (~30min) and I have 0.05%Errors. Is it possible that the average is calculated with the errors and 90% calculated without? Or can anybody eplain me why this happens?
regards
Stefan
--
Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!
Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org