You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@jmeter.apache.org by Katerina Novotna <ka...@redhat.com> on 2016/06/17 08:12:07 UTC

jmeter.reportgenerator.sample_filter usage

Hi, 

I'm trying to use jmeter.reportgenerator.sample_filter property to not to generate graph report from certain samples.

My .jtl log contains a lot of these samples:
1466145697314,78,G,200,OK,warm_group 1-1,text,true,,430,1,1,78,0
1466146126277,1,G,200,OK,stepping_group 2-454,text,true,,430,10,10,1,0

I set "jmeter.reportgenerator.sample_filter=^.*(stepping).*$" to generate graphs from samples from stepping_group not including warm_group.
But after this I get only empty graphs. 

What is the correct usage?

Regards,
Katka

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


Re: jmeter.reportgenerator.sample_filter usage + .jtl log size

Posted by Deepak Shetty <sh...@gmail.com>.
Hi
yes the enhancement was for things like responseMessage which i think the
dashboard should be able to handle without it being mandatory.(or disable
functionality where the data is not available without a hard error)

On Mon, Jun 20, 2016 at 8:56 AM, sebb <se...@gmail.com> wrote:

> On 20 June 2016 at 16:14, Deepak Shetty <sh...@gmail.com> wrote:
> >>I read from different sources that any listeners added into the thread
> > groups, utilize jmeter more,
> > True - but it may not be significant (and you cvan always add more JMeter
> > instances if it is)
> >
> >>1) To use generate dashboard report in jmeter3 I am forced to keep
> default
> > values of jmeter.save.saveservice.* properties
> > Yes - per
> >
> http://jmeter.apache.org/usermanual/generating-dashboard.html#configuration_requirements
>
> The link shows which settings are needed so that the dashboard has the
> data it needs.
>
> However you can change other saveservice parameters so as to omit their
> data.
>
> >>Is there any way to tell jmeter3 not to reguire this parameter (or any
> > other) for the report generation?
> > I dont think so - you probably need an enhancement request
>
> See above
>
> >>2) I'm struggling with the size of the .jtl file.
> > You can either
> > a. Process the file after your test runs to remove stuff you dont need.
> > b. Write custom listeners that can filter out what you dont need
> > c. Use a listener that can write data to stores for whom large data isnt
> an
> > issue at all (Backend listeners in JMeter or RDBMS)
> > I prefer c.
> >
> > On Mon, Jun 20, 2016 at 5:03 AM, Katerina Novotna <ka...@redhat.com>
> > wrote:
> >
> >> Thank you for reply. Please see my comments below.
> >>
> >> ----- Original Message -----
> >> > are you by any chance trying to filter on thread group name ? I
> believe
> >> the
> >> > sample filter only applies to Sample names - so you would have to
> append
> >> a
> >> > thread group name to the sampler name if you wanted to do this ? You
> can
> >> > also use any OS tool to filter the file before passing it on to
> JMeter.
> >> That did it:), sample_filter applies for sample name not the whole line.
> >>
> >> > An easier way if you want to filter an entire thread group is to make
> a
> >> > listener that is scoped only to the thread group you want (by
> creating it
> >> > under the the thread group rather than test plan)
> >> I read from different sources that any listeners added into the thread
> >> groups, utilize jmeter more, so I am using non-gui mode and logging
> with:
> >> ./jmeter -n -t my_test_plan.jmx -l jmeter-log.jtl (I specify the content
> >> of the .jtl file with jmeter.save.saveservice.* properties)
> >>
> >> This brings more questions:
> >> 1) To use generate dashboard report in jmeter3 I am forced to keep
> default
> >> values of jmeter.save.saveservice.* properties. Otherwise the generator
> >> says:
> >>
> >> "An error occurred: Error while processing samples:Consumer failed with
> >> message :Consumer failed with message :Consumer failed with message :No
> >> column <responseMessage> found in sample metadata
> >>
> <timeStamp,elapsed,label,responseCode,threadName,success,grpThreads,allThreads,Latency,IdleTime>,
> >> check #jmeter.save.saveservice.* properties to add the missing column"
> >>
> >> Is there any way to tell jmeter3 not to reguire this parameter (or any
> >> other) for the report generation?
> >>
> >> 2) I'm struggling with the size of the .jtl file. I tried to eliminate
> all
> >> the fields from .jtl file to not to log what I don't need. But still I
> get
> >> 1.3G of data per hour.
> >> I want to run soak tests for 3 days for example. I don't need to log
> every
> >> sample. Something like Summariser listener but applied to the samples. I
> >> couldn't find any jmeter plugin nor recommendation how to process this.
> >> Any advice would be very helpful.
> >>
> >>
> >>
> >>
> >>
> >> > On Fri, Jun 17, 2016 at 1:12 AM, Katerina Novotna <
> kanovotn@redhat.com>
> >> > wrote:
> >> >
> >> > > Hi,
> >> > >
> >> > > I'm trying to use jmeter.reportgenerator.sample_filter property to
> not
> >> to
> >> > > generate graph report from certain samples.
> >> > >
> >> > > My .jtl log contains a lot of these samples:
> >> > > 1466145697314,78,G,200,OK,warm_group 1-1,text,true,,430,1,1,78,0
> >> > > 1466146126277,1,G,200,OK,stepping_group
> 2-454,text,true,,430,10,10,1,0
> >> > >
> >> > > I set "jmeter.reportgenerator.sample_filter=^.*(stepping).*$" to
> >> generate
> >> > > graphs from samples from stepping_group not including warm_group.
> >> > > But after this I get only empty graphs.
> >> > >
> >> > > What is the correct usage?
> >> > >
> >> > > Regards,
> >> > > Katka
> >> > >
> >> > >
> ---------------------------------------------------------------------
> >> > > 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
> >>
> >>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
> For additional commands, e-mail: user-help@jmeter.apache.org
>
>

Re: jmeter.reportgenerator.sample_filter usage + .jtl log size

Posted by sebb <se...@gmail.com>.
On 20 June 2016 at 16:14, Deepak Shetty <sh...@gmail.com> wrote:
>>I read from different sources that any listeners added into the thread
> groups, utilize jmeter more,
> True - but it may not be significant (and you cvan always add more JMeter
> instances if it is)
>
>>1) To use generate dashboard report in jmeter3 I am forced to keep default
> values of jmeter.save.saveservice.* properties
> Yes - per
> http://jmeter.apache.org/usermanual/generating-dashboard.html#configuration_requirements

The link shows which settings are needed so that the dashboard has the
data it needs.

However you can change other saveservice parameters so as to omit their data.

>>Is there any way to tell jmeter3 not to reguire this parameter (or any
> other) for the report generation?
> I dont think so - you probably need an enhancement request

See above

>>2) I'm struggling with the size of the .jtl file.
> You can either
> a. Process the file after your test runs to remove stuff you dont need.
> b. Write custom listeners that can filter out what you dont need
> c. Use a listener that can write data to stores for whom large data isnt an
> issue at all (Backend listeners in JMeter or RDBMS)
> I prefer c.
>
> On Mon, Jun 20, 2016 at 5:03 AM, Katerina Novotna <ka...@redhat.com>
> wrote:
>
>> Thank you for reply. Please see my comments below.
>>
>> ----- Original Message -----
>> > are you by any chance trying to filter on thread group name ? I believe
>> the
>> > sample filter only applies to Sample names - so you would have to append
>> a
>> > thread group name to the sampler name if you wanted to do this ? You can
>> > also use any OS tool to filter the file before passing it on to JMeter.
>> That did it:), sample_filter applies for sample name not the whole line.
>>
>> > An easier way if you want to filter an entire thread group is to make a
>> > listener that is scoped only to the thread group you want (by creating it
>> > under the the thread group rather than test plan)
>> I read from different sources that any listeners added into the thread
>> groups, utilize jmeter more, so I am using non-gui mode and logging with:
>> ./jmeter -n -t my_test_plan.jmx -l jmeter-log.jtl (I specify the content
>> of the .jtl file with jmeter.save.saveservice.* properties)
>>
>> This brings more questions:
>> 1) To use generate dashboard report in jmeter3 I am forced to keep default
>> values of jmeter.save.saveservice.* properties. Otherwise the generator
>> says:
>>
>> "An error occurred: Error while processing samples:Consumer failed with
>> message :Consumer failed with message :Consumer failed with message :No
>> column <responseMessage> found in sample metadata
>> <timeStamp,elapsed,label,responseCode,threadName,success,grpThreads,allThreads,Latency,IdleTime>,
>> check #jmeter.save.saveservice.* properties to add the missing column"
>>
>> Is there any way to tell jmeter3 not to reguire this parameter (or any
>> other) for the report generation?
>>
>> 2) I'm struggling with the size of the .jtl file. I tried to eliminate all
>> the fields from .jtl file to not to log what I don't need. But still I get
>> 1.3G of data per hour.
>> I want to run soak tests for 3 days for example. I don't need to log every
>> sample. Something like Summariser listener but applied to the samples. I
>> couldn't find any jmeter plugin nor recommendation how to process this.
>> Any advice would be very helpful.
>>
>>
>>
>>
>>
>> > On Fri, Jun 17, 2016 at 1:12 AM, Katerina Novotna <ka...@redhat.com>
>> > wrote:
>> >
>> > > Hi,
>> > >
>> > > I'm trying to use jmeter.reportgenerator.sample_filter property to not
>> to
>> > > generate graph report from certain samples.
>> > >
>> > > My .jtl log contains a lot of these samples:
>> > > 1466145697314,78,G,200,OK,warm_group 1-1,text,true,,430,1,1,78,0
>> > > 1466146126277,1,G,200,OK,stepping_group 2-454,text,true,,430,10,10,1,0
>> > >
>> > > I set "jmeter.reportgenerator.sample_filter=^.*(stepping).*$" to
>> generate
>> > > graphs from samples from stepping_group not including warm_group.
>> > > But after this I get only empty graphs.
>> > >
>> > > What is the correct usage?
>> > >
>> > > Regards,
>> > > Katka
>> > >
>> > > ---------------------------------------------------------------------
>> > > 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
>>
>>

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


Re: jmeter.reportgenerator.sample_filter usage + .jtl log size

Posted by Deepak Shetty <sh...@gmail.com>.
>I read from different sources that any listeners added into the thread
groups, utilize jmeter more,
True - but it may not be significant (and you cvan always add more JMeter
instances if it is)

>1) To use generate dashboard report in jmeter3 I am forced to keep default
values of jmeter.save.saveservice.* properties
Yes - per
http://jmeter.apache.org/usermanual/generating-dashboard.html#configuration_requirements
>Is there any way to tell jmeter3 not to reguire this parameter (or any
other) for the report generation?
I dont think so - you probably need an enhancement request

>2) I'm struggling with the size of the .jtl file.
You can either
a. Process the file after your test runs to remove stuff you dont need.
b. Write custom listeners that can filter out what you dont need
c. Use a listener that can write data to stores for whom large data isnt an
issue at all (Backend listeners in JMeter or RDBMS)
I prefer c.


On Mon, Jun 20, 2016 at 5:03 AM, Katerina Novotna <ka...@redhat.com>
wrote:

> Thank you for reply. Please see my comments below.
>
> ----- Original Message -----
> > are you by any chance trying to filter on thread group name ? I believe
> the
> > sample filter only applies to Sample names - so you would have to append
> a
> > thread group name to the sampler name if you wanted to do this ? You can
> > also use any OS tool to filter the file before passing it on to JMeter.
> That did it:), sample_filter applies for sample name not the whole line.
>
> > An easier way if you want to filter an entire thread group is to make a
> > listener that is scoped only to the thread group you want (by creating it
> > under the the thread group rather than test plan)
> I read from different sources that any listeners added into the thread
> groups, utilize jmeter more, so I am using non-gui mode and logging with:
> ./jmeter -n -t my_test_plan.jmx -l jmeter-log.jtl (I specify the content
> of the .jtl file with jmeter.save.saveservice.* properties)
>
> This brings more questions:
> 1) To use generate dashboard report in jmeter3 I am forced to keep default
> values of jmeter.save.saveservice.* properties. Otherwise the generator
> says:
>
> "An error occurred: Error while processing samples:Consumer failed with
> message :Consumer failed with message :Consumer failed with message :No
> column <responseMessage> found in sample metadata
> <timeStamp,elapsed,label,responseCode,threadName,success,grpThreads,allThreads,Latency,IdleTime>,
> check #jmeter.save.saveservice.* properties to add the missing column"
>
> Is there any way to tell jmeter3 not to reguire this parameter (or any
> other) for the report generation?
>
> 2) I'm struggling with the size of the .jtl file. I tried to eliminate all
> the fields from .jtl file to not to log what I don't need. But still I get
> 1.3G of data per hour.
> I want to run soak tests for 3 days for example. I don't need to log every
> sample. Something like Summariser listener but applied to the samples. I
> couldn't find any jmeter plugin nor recommendation how to process this.
> Any advice would be very helpful.
>
>
>
>
>
> > On Fri, Jun 17, 2016 at 1:12 AM, Katerina Novotna <ka...@redhat.com>
> > wrote:
> >
> > > Hi,
> > >
> > > I'm trying to use jmeter.reportgenerator.sample_filter property to not
> to
> > > generate graph report from certain samples.
> > >
> > > My .jtl log contains a lot of these samples:
> > > 1466145697314,78,G,200,OK,warm_group 1-1,text,true,,430,1,1,78,0
> > > 1466146126277,1,G,200,OK,stepping_group 2-454,text,true,,430,10,10,1,0
> > >
> > > I set "jmeter.reportgenerator.sample_filter=^.*(stepping).*$" to
> generate
> > > graphs from samples from stepping_group not including warm_group.
> > > But after this I get only empty graphs.
> > >
> > > What is the correct usage?
> > >
> > > Regards,
> > > Katka
> > >
> > > ---------------------------------------------------------------------
> > > 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: jmeter.reportgenerator.sample_filter usage + .jtl log size

Posted by Katerina Novotna <ka...@redhat.com>.
Thank you for reply. Please see my comments below.

----- Original Message -----
> are you by any chance trying to filter on thread group name ? I believe the
> sample filter only applies to Sample names - so you would have to append a
> thread group name to the sampler name if you wanted to do this ? You can
> also use any OS tool to filter the file before passing it on to JMeter.
That did it:), sample_filter applies for sample name not the whole line.

> An easier way if you want to filter an entire thread group is to make a
> listener that is scoped only to the thread group you want (by creating it
> under the the thread group rather than test plan)
I read from different sources that any listeners added into the thread groups, utilize jmeter more, so I am using non-gui mode and logging with:
./jmeter -n -t my_test_plan.jmx -l jmeter-log.jtl (I specify the content of the .jtl file with jmeter.save.saveservice.* properties)

This brings more questions:
1) To use generate dashboard report in jmeter3 I am forced to keep default values of jmeter.save.saveservice.* properties. Otherwise the generator says:

"An error occurred: Error while processing samples:Consumer failed with message :Consumer failed with message :Consumer failed with message :No column <responseMessage> found in sample metadata <timeStamp,elapsed,label,responseCode,threadName,success,grpThreads,allThreads,Latency,IdleTime>, check #jmeter.save.saveservice.* properties to add the missing column"

Is there any way to tell jmeter3 not to reguire this parameter (or any other) for the report generation?

2) I'm struggling with the size of the .jtl file. I tried to eliminate all the fields from .jtl file to not to log what I don't need. But still I get 1.3G of data per hour.
I want to run soak tests for 3 days for example. I don't need to log every sample. Something like Summariser listener but applied to the samples. I couldn't find any jmeter plugin nor recommendation how to process this.
Any advice would be very helpful. 





> On Fri, Jun 17, 2016 at 1:12 AM, Katerina Novotna <ka...@redhat.com>
> wrote:
> 
> > Hi,
> >
> > I'm trying to use jmeter.reportgenerator.sample_filter property to not to
> > generate graph report from certain samples.
> >
> > My .jtl log contains a lot of these samples:
> > 1466145697314,78,G,200,OK,warm_group 1-1,text,true,,430,1,1,78,0
> > 1466146126277,1,G,200,OK,stepping_group 2-454,text,true,,430,10,10,1,0
> >
> > I set "jmeter.reportgenerator.sample_filter=^.*(stepping).*$" to generate
> > graphs from samples from stepping_group not including warm_group.
> > But after this I get only empty graphs.
> >
> > What is the correct usage?
> >
> > Regards,
> > Katka
> >
> > ---------------------------------------------------------------------
> > 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: jmeter.reportgenerator.sample_filter usage

Posted by Deepak Shetty <sh...@gmail.com>.
are you by any chance trying to filter on thread group name ? I believe the
sample filter only applies to Sample names - so you would have to append a
thread group name to the sampler name if you wanted to do this ? You can
also use any OS tool to filter the file before passing it on to JMeter.
An easier way if you want to filter an entire thread group is to make a
listener that is scoped only to the thread group you want (by creating it
under the the thread group rather than test plan)

On Fri, Jun 17, 2016 at 1:12 AM, Katerina Novotna <ka...@redhat.com>
wrote:

> Hi,
>
> I'm trying to use jmeter.reportgenerator.sample_filter property to not to
> generate graph report from certain samples.
>
> My .jtl log contains a lot of these samples:
> 1466145697314,78,G,200,OK,warm_group 1-1,text,true,,430,1,1,78,0
> 1466146126277,1,G,200,OK,stepping_group 2-454,text,true,,430,10,10,1,0
>
> I set "jmeter.reportgenerator.sample_filter=^.*(stepping).*$" to generate
> graphs from samples from stepping_group not including warm_group.
> But after this I get only empty graphs.
>
> What is the correct usage?
>
> Regards,
> Katka
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@jmeter.apache.org
> For additional commands, e-mail: user-help@jmeter.apache.org
>
>