You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@jmeter.apache.org by Pieter Ennes <pi...@watchmouse.com> on 2009/07/31 14:24:30 UTC

Summing attributes of child elements

Hello,

We're a bit confused about the numbers in this XML snippet (edited for 
readability, hope not too much removed):

<httpSample t="3773" lt="0" [...] rc="200" rm="OK" by="709260">
   <httpSample t="2608" lt="595" rc="302" rm="Found"
by="321851"/>
   <httpSample t="1163" lt="988" rc="200" rm="OK" by="65558"/>
</httpSample>

Element had follow redirects on, redirect automatically off.

t  =   2608 +  1163 =   3771
lt =    595 +   988 =   1583
by = 321851 + 65558 = 387409

Now the parent element has quite different numbers for all three 
attributes, and also in a different way.

- 't' has just 2 bytes too many
- 'lt' I would expect the parent to have 1583 (sum) or 595 (first)
- 'by' is off by almost a factor of 2

If this is the way the math for the parent element is meant to work, is 
it documented somewhere?

Thanks,
-- 
  - Pieter

---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org


Re: Summing attributes of child elements

Posted by sebb <se...@gmail.com>.
On 31/07/2009, Pieter Ennes <pi...@watchmouse.com> wrote:
> On 31-07-09 13:24, Pieter Ennes wrote:
>
> > Hello,
> >
> > We're a bit confused about the numbers in this XML snippet (edited for
> > readability, hope not too much removed):
> >
> > <httpSample t="3773" lt="0" [...] rc="200" rm="OK" by="709260">
> > <httpSample t="2608" lt="595" rc="302" rm="Found"
> > by="321851"/>

Seems like a very large size for a redirect.

> > <httpSample t="1163" lt="988" rc="200" rm="OK" by="65558"/>
> > </httpSample>
> >
> > Element had follow redirects on, redirect automatically off.
> >
> > t = 2608 + 1163 = 3771

There may be slight gaps between the samples so the overall time can
be slightly longer than the sum of the individual samples.

> > lt = 595 + 988 = 1583

Latency cannot be added; it applies to a single sample only.

> > by = 321851 + 65558 = 387409
> >
>
>  Heh, my colleague found out that for 'by' coincidentally:
>
>  Total (709260) - Expected (387409) is exactly the size of the first request
> (321851)
>
>  Which is a bit odd at least, and would suggest that the byte size of first
> request is counted twice in the parent. Not sure how this translates to the
> timing though.

Yes, that does look a bit odd; need to look further at that.

>  --
>   - Pieter
>
> ---------------------------------------------------------------------
>  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


Re: Summing attributes of child elements

Posted by Pieter Ennes <pi...@watchmouse.com>.
On 31-07-09 13:24, Pieter Ennes wrote:
> Hello,
>
> We're a bit confused about the numbers in this XML snippet (edited for
> readability, hope not too much removed):
>
> <httpSample t="3773" lt="0" [...] rc="200" rm="OK" by="709260">
> <httpSample t="2608" lt="595" rc="302" rm="Found"
> by="321851"/>
> <httpSample t="1163" lt="988" rc="200" rm="OK" by="65558"/>
> </httpSample>
>
> Element had follow redirects on, redirect automatically off.
>
> t = 2608 + 1163 = 3771
> lt = 595 + 988 = 1583
> by = 321851 + 65558 = 387409

Heh, my colleague found out that for 'by' coincidentally:

Total (709260) - Expected (387409) is exactly the size of the first 
request (321851)

Which is a bit odd at least, and would suggest that the byte size of 
first request is counted twice in the parent. Not sure how this 
translates to the timing though.
-- 
  - Pieter

---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org