You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@jmeter.apache.org by Milamber <mi...@apache.org> on 2016/10/19 18:13:42 UTC
Re: svn commit: r1765062 - in /jmeter/trunk:
src/components/org/apache/jmeter/visualizers/
src/core/org/apache/jmeter/resources/ src/core/org/apache/jmeter/util/
src/core/org/apache/jmeter/visualizers/ src/jorphan/org/apache/jorphan/math/
xdocs/ xdocs/user...
Philippe,
Probably a regression with the getRate() method?
Before :
return ((double) count / (double) elapsedTime ) * 1000;
After:
return value / ((double) elapsedTime / 1000); // 1000 = millisecs/sec
* 1000 --> / 1000
On 15/10/2016 14:34, pmouawad@apache.org wrote:
> Modified: jmeter/trunk/src/core/org/apache/jmeter/util/Calculator.java
> URL:http://svn.apache.org/viewvc/jmeter/trunk/src/core/org/apache/jmeter/util/Calculator.java?rev=1765062&r1=1765061&r2=1765062&view=diff
> ==============================================================================
> --- jmeter/trunk/src/core/org/apache/jmeter/util/Calculator.java (original)
> +++ jmeter/trunk/src/core/org/apache/jmeter/util/Calculator.java Sat Oct 15 13:34:54 2016
[snip]
> public long getTotalBytes() {
> return bytes;
> @@ -181,11 +200,7 @@ public class Calculator {
> * @return throughput associated to this sampler in requests per second
> */
> public double getRate() {
> - if (elapsedTime == 0) {
> - return 0.0;
> - }
> -
> - return ((double) count / (double) elapsedTime ) * 1000;
> + return getRatePerSecond(count);
> }
>
[snip]
> + /**
> + *
> + * @param value value for which we compute rate
> + * @return double rate
> + */
> + private double getRatePerSecond(long value) {
> + if (elapsedTime > 0) {
> + return value / ((double) elapsedTime / 1000); // 1000 = millisecs/sec
> + }
> + return 0.0;
> + }
>
> }
Re: svn commit: r1765062 - in /jmeter/trunk:
src/components/org/apache/jmeter/visualizers/
src/core/org/apache/jmeter/resources/ src/core/org/apache/jmeter/util/
src/core/org/apache/jmeter/visualizers/ src/jorphan/org/apache/jorphan/math/
xdocs/ xdocs/user...
Posted by Milamber <mi...@apache.org>.
You are right. I have not seen the parentheses...
On 19/10/2016 20:55, Philippe Mouawad wrote:
> Hi Milamber,
> I don't think so:
>
> ((double) value / (double) elapsedTime ) * 1000 == value / ((double)
> elapsedTime / 1000)
>
> 1/1/1000 = *1000
>
> Unless I misunderstand the problem.
> Thanks
>
> On Wed, Oct 19, 2016 at 8:13 PM, Milamber <mi...@apache.org> wrote:
>
>> Philippe,
>>
>> Probably a regression with the getRate() method?
>>
>> Before :
>> return ((double) count / (double) elapsedTime ) * 1000;
>>
>> After:
>>
>> return value / ((double) elapsedTime / 1000); // 1000 = millisecs/sec
>>
>> * 1000 --> / 1000
>>
>>
>>
>> On 15/10/2016 14:34, pmouawad@apache.org wrote:
>>
>>> Modified: jmeter/trunk/src/core/org/apache/jmeter/util/Calculator.java
>>> URL:http://svn.apache.org/viewvc/jmeter/trunk/src/core/org/
>>> apache/jmeter/util/Calculator.java?rev=1765062&r1=1765061&
>>> r2=1765062&view=diff
>>> ============================================================
>>> ==================
>>> --- jmeter/trunk/src/core/org/apache/jmeter/util/Calculator.java
>>> (original)
>>> +++ jmeter/trunk/src/core/org/apache/jmeter/util/Calculator.java Sat Oct
>>> 15 13:34:54 2016
>>>
>> [snip]
>>
>>
>>
>> public long getTotalBytes() {
>>> return bytes;
>>> @@ -181,11 +200,7 @@ public class Calculator {
>>> * @return throughput associated to this sampler in requests per
>>> second
>>> */
>>> public double getRate() {
>>> - if (elapsedTime == 0) {
>>> - return 0.0;
>>> - }
>>> -
>>> - return ((double) count / (double) elapsedTime ) * 1000;
>>> + return getRatePerSecond(count);
>>> }
>>>
>>>
>> [snip]
>>
>>
>> + /**
>>> + *
>>> + * @param value value for which we compute rate
>>> + * @return double rate
>>> + */
>>> + private double getRatePerSecond(long value) {
>>> + if (elapsedTime > 0) {
>>> + return value / ((double) elapsedTime / 1000); // 1000 =
>>> millisecs/sec
>>> + }
>>> + return 0.0;
>>> + }
>>> }
>>>
>>
>
Re: svn commit: r1765062 - in /jmeter/trunk: src/components/org/apache/jmeter/visualizers/
src/core/org/apache/jmeter/resources/ src/core/org/apache/jmeter/util/
src/core/org/apache/jmeter/visualizers/ src/jorphan/org/apache/jorphan/math/
xdocs/ xdocs/user...
Posted by Philippe Mouawad <ph...@gmail.com>.
Hi Milamber,
I don't think so:
((double) value / (double) elapsedTime ) * 1000 == value / ((double)
elapsedTime / 1000)
1/1/1000 = *1000
Unless I misunderstand the problem.
Thanks
On Wed, Oct 19, 2016 at 8:13 PM, Milamber <mi...@apache.org> wrote:
> Philippe,
>
> Probably a regression with the getRate() method?
>
> Before :
> return ((double) count / (double) elapsedTime ) * 1000;
>
> After:
>
> return value / ((double) elapsedTime / 1000); // 1000 = millisecs/sec
>
> * 1000 --> / 1000
>
>
>
> On 15/10/2016 14:34, pmouawad@apache.org wrote:
>
>> Modified: jmeter/trunk/src/core/org/apache/jmeter/util/Calculator.java
>> URL:http://svn.apache.org/viewvc/jmeter/trunk/src/core/org/
>> apache/jmeter/util/Calculator.java?rev=1765062&r1=1765061&
>> r2=1765062&view=diff
>> ============================================================
>> ==================
>> --- jmeter/trunk/src/core/org/apache/jmeter/util/Calculator.java
>> (original)
>> +++ jmeter/trunk/src/core/org/apache/jmeter/util/Calculator.java Sat Oct
>> 15 13:34:54 2016
>>
>
> [snip]
>
>
>
> public long getTotalBytes() {
>> return bytes;
>> @@ -181,11 +200,7 @@ public class Calculator {
>> * @return throughput associated to this sampler in requests per
>> second
>> */
>> public double getRate() {
>> - if (elapsedTime == 0) {
>> - return 0.0;
>> - }
>> -
>> - return ((double) count / (double) elapsedTime ) * 1000;
>> + return getRatePerSecond(count);
>> }
>>
>>
> [snip]
>
>
> + /**
>> + *
>> + * @param value value for which we compute rate
>> + * @return double rate
>> + */
>> + private double getRatePerSecond(long value) {
>> + if (elapsedTime > 0) {
>> + return value / ((double) elapsedTime / 1000); // 1000 =
>> millisecs/sec
>> + }
>> + return 0.0;
>> + }
>> }
>>
>
>
--
Cordialement.
Philippe Mouawad.