Performance Test Setup for a web site...

[Harry_]

Hi,

We need to do performance testing for our website simulating 5000 users using jmeter and other open source tools. The following information is with us:

- A csv file containing links and number of times that link last month. The file is sorted according to popularity of page visited.
- With this file I can get information about average number of hits per unit time say per minute. (estimate can be made regarding max load, min load). I will create Scripts which will hit the site pages proportional to there usage.
- All the requests will be http requests.
- Average size of the page will be 350 KB (including embedded objects within the page).

Now we need answer to the following questions:

- How should performance testing of the site simulating that much user load using Jmeter should be done?
- What should be the configuration of the machine which will simulate these many users? (Someone suggested me 8 core@3 GHz, 16 GB machine).
- If more then one test machine is required please specify there configuration?
- How many instances of jMeter we need to run for simulating 5000 users?
- How the result file output should be consolidated from various scripts? If someone is using any tool for consolidating the output files please share information regarding the same.
- Is there any tool for converting these files to some reports. I have heard that reports can be generated using some available xslt's in ant. But that does not work with large output files. So please suggest alternatives.
- What other things should i keep in mind for doing the performance test?

We would appreciate if someone can answer these queries based on there experience. .

Thanks,
Harry

[Deepak Shetty]

>I will create Scripts which will hit the site pages proportional to there
usage
This isnt easy.

>- What should be the configuration of the machine which will simulate these
>- If more then one test machine is required please specify there
configuration?
>- How many instances of jMeter we need to run for simulating 5000 users?
My preference is Multiple low end machines running separate JMeter instances
to 1 big machine. It simulates the network better . The load you can
generate depends on what else is running and what your tests actually do.
People have reported running 1000 threads from a single machine. In any case
, you must generate a load and check your client machine resources ,
preferably the cpu shouldn't exceed 60-80% and memory used should all be RAM
not virtual. You can increase the number of threads till you hit some limit
after which the client machine may become a bottleneck and give you
incorrect results.
See related
http://wiki.apache.org/jakarta-jmeter/HowManyThreads
http://jakarta.apache.org/jmeter/usermanual/best-practices.html#lean_mean
When running multiple jmeter instances you can either run each instance
separately (my preference) or you can run Jmeter  in master slave (this is
more inefficient , check the mail archives).

>How the result file output should be consolidated from various scripts? If
>someone is using any tool for consolidating the output files please share
>information regarding the same.
>- Is there any tool for converting these files to some reports. I have
heard
>that reports can be generated using some available xslt's in ant. But that
>does not work with large output files. So please suggest alternatives.
I assume you mean combining the results if you run jmeter instances
separately. if you use CSV as your format, just concatenate. Mostly if you
have long running tests , you would load the CSV files into a database
table. You have some listeners that can read the CSV/JTL files , but you'd
have to use a listener whose memory utlization is constant and not
proportional  to number of samples (e.g. summary report -
http://jakarta.apache.org/jmeter/usermanual/component_reference.html section
18.3).

regards
deepak


On Thu, Nov 5, 2009 at 2:54 PM, Harry_ <harjitworks@...> wrote:

>
> Hi,
>
> We need to do performance testing for our website simulating 5000 users
> using jmeter and other open source tools. The following information is with
> us:
>
> - A csv file containing links and number of times that link last month. The
> file is sorted according to popularity of page visited.
> - With this file I can get information about average number of hits per
> unit
> time say per minute. (estimate can be made regarding max load, min load). I
> will create Scripts which will hit the site pages proportional to there
> usage.
> - All the requests will be http requests.
> - Average size of the page will be 350 KB (including embedded objects
> within
> the page).
>
> Now we need answer to the following questions:
>
> - How should performance testing of the site simulating that much user load
> using Jmeter should be done?
> - What should be the configuration of the machine which will simulate these
> many users? (Someone suggested me 8 core@3 GHz, 16 GB machine).
> - If more then one test machine is required please specify there
> configuration?
> - How many instances of jMeter we need to run for simulating 5000 users?
> - How the result file output should be consolidated from various scripts?
> If
> someone is using any tool for consolidating the output files please share
> information regarding the same.
> - Is there any tool for converting these files to some reports. I have
> heard
> that reports can be generated using some available xslt's in ant. But that
> does not work with large output files. So please suggest alternatives.
> - What other things should i keep in mind for doing the performance test?
>
> We would appreciate if someone can answer these queries based on there
> experience. .
>
> Thanks,
> Harry
> --
> View this message in context:
> http://old.nabble.com/Performance-Test-Setup-for-a-web-site...-tp26223743p26223743.html
> Sent from the JMeter - User mailing list archive at Nabble.com.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@...
> For additional commands, e-mail: jmeter-user-help@...
>
>

[Carl Shaulis]

Getting proportional scripts I can think of two approaches:
1.  Scrap a days worth of log files for all of the requests coming to your
application servers then replay these requests using JMeter
2.  Estimate what portion of your community does what, then use various
thread groups to emulate this traffic (Thread group one (ie-visit home page)
50%, ThreadGroup2 (ie. - Search) 40% etc...

Number of virtual threads really is dependent on a combination of your load
servers and your target.  For example we have a 16 core machine as a load
generator and a an equivalent machine supporting our application.  Using 100
concurrent threads and no sleeps the load machine was less than 10% CPU
utilization and the target machine was at 90% CPU utilization.  When then
cached a bunch of the requests and the target machine could respond faster
than the load machine so the Load machine CPU was stressed at the same
thread count, additionally we were almost at our bandwidth limits.  So
Deepak is spot on saying monitor machine resources.

We have had success using MySQL to inspect data, but we have also kept our
load times brief (5 minutes).  If using a linux box to collect your data you
can get a quick evaluation of throughput by using wc -l <filename> divided
by the test duration to give you TPS.

Good luck!

Carl


On 11/6/09 1:13 AM, "Deepak Shetty" <shettyd@...> wrote:

>> I will create Scripts which will hit the site pages proportional to there
> usage
> This isnt easy.
>
>> - What should be the configuration of the machine which will simulate these
>> - If more then one test machine is required please specify there
> configuration?
>> - How many instances of jMeter we need to run for simulating 5000 users?
> My preference is Multiple low end machines running separate JMeter instances
> to 1 big machine. It simulates the network better . The load you can
> generate depends on what else is running and what your tests actually do.
> People have reported running 1000 threads from a single machine. In any case
> , you must generate a load and check your client machine resources ,
> preferably the cpu shouldn't exceed 60-80% and memory used should all be RAM
> not virtual. You can increase the number of threads till you hit some limit
> after which the client machine may become a bottleneck and give you
> incorrect results.
> See related
> http://wiki.apache.org/jakarta-jmeter/HowManyThreads
> http://jakarta.apache.org/jmeter/usermanual/best-practices.html#lean_mean
> When running multiple jmeter instances you can either run each instance
> separately (my preference) or you can run Jmeter  in master slave (this is
> more inefficient , check the mail archives).
>
>> How the result file output should be consolidated from various scripts? If
>> someone is using any tool for consolidating the output files please share
>> information regarding the same.
>> - Is there any tool for converting these files to some reports. I have
> heard
>> that reports can be generated using some available xslt's in ant. But that
>> does not work with large output files. So please suggest alternatives.
> I assume you mean combining the results if you run jmeter instances
> separately. if you use CSV as your format, just concatenate. Mostly if you
> have long running tests , you would load the CSV files into a database
> table. You have some listeners that can read the CSV/JTL files , but you'd
> have to use a listener whose memory utlization is constant and not
> proportional  to number of samples (e.g. summary report -
> http://jakarta.apache.org/jmeter/usermanual/component_reference.html section
> 18.3).
>
> regards
> deepak
>
>
> On Thu, Nov 5, 2009 at 2:54 PM, Harry_ <harjitworks@...> wrote:
>
>>
>> Hi,
>>
>> We need to do performance testing for our website simulating 5000 users
>> using jmeter and other open source tools. The following information is with
>> us:
>>
>> - A csv file containing links and number of times that link last month. The
>> file is sorted according to popularity of page visited.
>> - With this file I can get information about average number of hits per
>> unit
>> time say per minute. (estimate can be made regarding max load, min load). I
>> will create Scripts which will hit the site pages proportional to there
>> usage.
>> - All the requests will be http requests.
>> - Average size of the page will be 350 KB (including embedded objects
>> within
>> the page).
>>
>> Now we need answer to the following questions:
>>
>> - How should performance testing of the site simulating that much user load
>> using Jmeter should be done?
>> - What should be the configuration of the machine which will simulate these
>> many users? (Someone suggested me 8 core@3 GHz, 16 GB machine).
>> - If more then one test machine is required please specify there
>> configuration?
>> - How many instances of jMeter we need to run for simulating 5000 users?
>> - How the result file output should be consolidated from various scripts?
>> If
>> someone is using any tool for consolidating the output files please share
>> information regarding the same.
>> - Is there any tool for converting these files to some reports. I have
>> heard
>> that reports can be generated using some available xslt's in ant. But that
>> does not work with large output files. So please suggest alternatives.
>> - What other things should i keep in mind for doing the performance test?
>>
>> We would appreciate if someone can answer these queries based on there
>> experience. .
>>
>> Thanks,
>> Harry
>> --
>> View this message in context:
>> http://old.nabble.com/Performance-Test-Setup-for-a-web-site...-tp26223743p262
>> 23743.html
>> Sent from the JMeter - User mailing list archive at Nabble.com.
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: jmeter-user-unsubscribe@...
>> For additional commands, e-mail: jmeter-user-help@...
>>
>>

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

[sebb-2-2]

On 06/11/2009, Carl Shaulis <cshaulis@...> wrote:

> Getting proportional scripts I can think of two approaches:
>  1.  Scrap a days worth of log files for all of the requests coming to your
>  application servers then replay these requests using JMeter
>  2.  Estimate what portion of your community does what, then use various
>  thread groups to emulate this traffic (Thread group one (ie-visit home page)
>  50%, ThreadGroup2 (ie. - Search) 40% etc...
>
>  Number of virtual threads really is dependent on a combination of your load
>  servers and your target.  For example we have a 16 core machine as a load
>  generator and a an equivalent machine supporting our application.  Using 100
>  concurrent threads and no sleeps the load machine was less than 10% CPU
>  utilization and the target machine was at 90% CPU utilization.  When then
>  cached a bunch of the requests and the target machine could respond faster
>  than the load machine so the Load machine CPU was stressed at the same
>  thread count, additionally we were almost at our bandwidth limits.  So
>  Deepak is spot on saying monitor machine resources.
>
>  We have had success using MySQL to inspect data, but we have also kept our
>  load times brief (5 minutes).  If using a linux box to collect your data you
>  can get a quick evaluation of throughput by using wc -l <filename> divided
>  by the test duration to give you TPS.

You can also use

http://jakarta.apache.org/jmeter/usermanual/component_reference.html#Generate_Summary_Results

==

Once you have got a test plan working, you can compare the server logs
with the originals to check whether the load is close to what you want
to simulate, and adjust accordingly.

>  Good luck!
>
>
>  Carl
>
>
>
>  On 11/6/09 1:13 AM, "Deepak Shetty" <shettyd@...> wrote:
>
>  >> I will create Scripts which will hit the site pages proportional to there
>  > usage
>  > This isnt easy.
>  >
>  >> - What should be the configuration of the machine which will simulate these
>  >> - If more then one test machine is required please specify there
>  > configuration?
>  >> - How many instances of jMeter we need to run for simulating 5000 users?
>  > My preference is Multiple low end machines running separate JMeter instances
>  > to 1 big machine. It simulates the network better . The load you can
>  > generate depends on what else is running and what your tests actually do.
>  > People have reported running 1000 threads from a single machine. In any case
>  > , you must generate a load and check your client machine resources ,
>  > preferably the cpu shouldn't exceed 60-80% and memory used should all be RAM
>  > not virtual. You can increase the number of threads till you hit some limit
>  > after which the client machine may become a bottleneck and give you
>  > incorrect results.
>  > See related
>  > http://wiki.apache.org/jakarta-jmeter/HowManyThreads
>  > http://jakarta.apache.org/jmeter/usermanual/best-practices.html#lean_mean
>  > When running multiple jmeter instances you can either run each instance
>  > separately (my preference) or you can run Jmeter  in master slave (this is
>  > more inefficient , check the mail archives).
>  >
>  >> How the result file output should be consolidated from various scripts? If
>  >> someone is using any tool for consolidating the output files please share
>  >> information regarding the same.
>  >> - Is there any tool for converting these files to some reports. I have
>  > heard
>  >> that reports can be generated using some available xslt's in ant. But that
>  >> does not work with large output files. So please suggest alternatives.
>  > I assume you mean combining the results if you run jmeter instances
>  > separately. if you use CSV as your format, just concatenate. Mostly if you
>  > have long running tests , you would load the CSV files into a database
>  > table. You have some listeners that can read the CSV/JTL files , but you'd
>  > have to use a listener whose memory utlization is constant and not
>  > proportional  to number of samples (e.g. summary report -
>  > http://jakarta.apache.org/jmeter/usermanual/component_reference.html section
>  > 18.3).
>  >
>  > regards
>  > deepak
>  >
>  >
>  > On Thu, Nov 5, 2009 at 2:54 PM, Harry_ <harjitworks@...> wrote:
>  >
>  >>
>  >> Hi,
>  >>
>  >> We need to do performance testing for our website simulating 5000 users
>  >> using jmeter and other open source tools. The following information is with
>  >> us:
>  >>
>  >> - A csv file containing links and number of times that link last month. The
>  >> file is sorted according to popularity of page visited.
>  >> - With this file I can get information about average number of hits per
>  >> unit
>  >> time say per minute. (estimate can be made regarding max load, min load). I
>  >> will create Scripts which will hit the site pages proportional to there
>  >> usage.
>  >> - All the requests will be http requests.
>  >> - Average size of the page will be 350 KB (including embedded objects
>  >> within
>  >> the page).
>  >>
>  >> Now we need answer to the following questions:
>  >>
>  >> - How should performance testing of the site simulating that much user load
>  >> using Jmeter should be done?
>  >> - What should be the configuration of the machine which will simulate these
>  >> many users? (Someone suggested me 8 core@3 GHz, 16 GB machine).
>  >> - If more then one test machine is required please specify there
>  >> configuration?
>  >> - How many instances of jMeter we need to run for simulating 5000 users?
>  >> - How the result file output should be consolidated from various scripts?
>  >> If
>  >> someone is using any tool for consolidating the output files please share
>  >> information regarding the same.
>  >> - Is there any tool for converting these files to some reports. I have
>  >> heard
>  >> that reports can be generated using some available xslt's in ant. But that
>  >> does not work with large output files. So please suggest alternatives.
>  >> - What other things should i keep in mind for doing the performance test?
>  >>
>  >> We would appreciate if someone can answer these queries based on there
>  >> experience. .
>  >>
>  >> Thanks,
>  >> Harry
>  >> --
>  >> View this message in context:
>  >> http://old.nabble.com/Performance-Test-Setup-for-a-web-site...-tp26223743p262
>  >> 23743.html
>  >> Sent from the JMeter - User mailing list archive at Nabble.com.
>  >>
>  >>
>  >> ---------------------------------------------------------------------
>  >> To unsubscribe, e-mail: jmeter-user-unsubscribe@...
>  >> For additional commands, e-mail: jmeter-user-help@...
>  >>
>  >>
>
>
>  ---------------------------------------------------------------------
>  To unsubscribe, e-mail: jmeter-user-unsubscribe@...
>  For additional commands, e-mail: jmeter-user-help@...
>
>

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

[Deepak Shetty]

Hi
one of the key problems that I face is simulating varying concurrent usage
accurately. e.g. when the day starts out you want this to be low and hit
peaks as people come into work and then drop off at lunch or whatever.
Specifying different thread groups is one of the ways that you simulate this
as you have pointed out.
JMeter doesn't have a way to vary thread counts in a thread group easily ,
you must setup the total number a priori and then use ramp ups or delays or
whatever.
Is there are any plans to be able to vary the number of active threads in
the threadgroup at runtime?
regards
deepak

On Fri, Nov 6, 2009 at 6:54 AM, Carl Shaulis <cshaulis@...> wrote:

> Getting proportional scripts I can think of two approaches:
> 1.  Scrap a days worth of log files for all of the requests coming to your
> application servers then replay these requests using JMeter
> 2.  Estimate what portion of your community does what, then use various
> thread groups to emulate this traffic (Thread group one (ie-visit home
> page)
> 50%, ThreadGroup2 (ie. - Search) 40% etc...
>
> Number of virtual threads really is dependent on a combination of your load
> servers and your target.  For example we have a 16 core machine as a load
> generator and a an equivalent machine supporting our application.  Using
> 100
> concurrent threads and no sleeps the load machine was less than 10% CPU
> utilization and the target machine was at 90% CPU utilization.  When then
> cached a bunch of the requests and the target machine could respond faster
> than the load machine so the Load machine CPU was stressed at the same
> thread count, additionally we were almost at our bandwidth limits.  So
> Deepak is spot on saying monitor machine resources.
>
> We have had success using MySQL to inspect data, but we have also kept our
> load times brief (5 minutes).  If using a linux box to collect your data
> you
> can get a quick evaluation of throughput by using wc -l <filename> divided
> by the test duration to give you TPS.
>
> Good luck!
>
> Carl
>
>
> On 11/6/09 1:13 AM, "Deepak Shetty" <shettyd@...> wrote:
>
> >> I will create Scripts which will hit the site pages proportional to
> there
> > usage
> > This isnt easy.
> >
> >> - What should be the configuration of the machine which will simulate
> these
> >> - If more then one test machine is required please specify there
> > configuration?
> >> - How many instances of jMeter we need to run for simulating 5000 users?
> > My preference is Multiple low end machines running separate JMeter
> instances
> > to 1 big machine. It simulates the network better . The load you can
> > generate depends on what else is running and what your tests actually do.
> > People have reported running 1000 threads from a single machine. In any
> case
> > , you must generate a load and check your client machine resources ,
> > preferably the cpu shouldn't exceed 60-80% and memory used should all be
> RAM
> > not virtual. You can increase the number of threads till you hit some
> limit
> > after which the client machine may become a bottleneck and give you
> > incorrect results.
> > See related
> > http://wiki.apache.org/jakarta-jmeter/HowManyThreads
> >
> http://jakarta.apache.org/jmeter/usermanual/best-practices.html#lean_mean
> > When running multiple jmeter instances you can either run each instance
> > separately (my preference) or you can run Jmeter  in master slave (this
> is
> > more inefficient , check the mail archives).
> >
> >> How the result file output should be consolidated from various scripts?
> If
> >> someone is using any tool for consolidating the output files please
> share
> >> information regarding the same.
> >> - Is there any tool for converting these files to some reports. I have
> > heard
> >> that reports can be generated using some available xslt's in ant. But
> that
> >> does not work with large output files. So please suggest alternatives.
> > I assume you mean combining the results if you run jmeter instances
> > separately. if you use CSV as your format, just concatenate. Mostly if
> you
> > have long running tests , you would load the CSV files into a database
> > table. You have some listeners that can read the CSV/JTL files , but
> you'd
> > have to use a listener whose memory utlization is constant and not
> > proportional  to number of samples (e.g. summary report -
> > http://jakarta.apache.org/jmeter/usermanual/component_reference.htmlsection
> > 18.3).
> >
> > regards
> > deepak
> >
> >
> > On Thu, Nov 5, 2009 at 2:54 PM, Harry_ <harjitworks@...> wrote:
> >
> >>
> >> Hi,
> >>
> >> We need to do performance testing for our website simulating 5000 users
> >> using jmeter and other open source tools. The following information is
> with
> >> us:
> >>
> >> - A csv file containing links and number of times that link last month.
> The
> >> file is sorted according to popularity of page visited.
> >> - With this file I can get information about average number of hits per
> >> unit
> >> time say per minute. (estimate can be made regarding max load, min
> load). I
> >> will create Scripts which will hit the site pages proportional to there
> >> usage.
> >> - All the requests will be http requests.
> >> - Average size of the page will be 350 KB (including embedded objects
> >> within
> >> the page).
> >>
> >> Now we need answer to the following questions:
> >>
> >> - How should performance testing of the site simulating that much user
> load
> >> using Jmeter should be done?
> >> - What should be the configuration of the machine which will simulate
> these
> >> many users? (Someone suggested me 8 core@3 GHz, 16 GB machine).
> >> - If more then one test machine is required please specify there
> >> configuration?
> >> - How many instances of jMeter we need to run for simulating 5000 users?
> >> - How the result file output should be consolidated from various
> scripts?
> >> If
> >> someone is using any tool for consolidating the output files please
> share
> >> information regarding the same.
> >> - Is there any tool for converting these files to some reports. I have
> >> heard
> >> that reports can be generated using some available xslt's in ant. But
> that
> >> does not work with large output files. So please suggest alternatives.
> >> - What other things should i keep in mind for doing the performance
> test?
> >>
> >> We would appreciate if someone can answer these queries based on there
> >> experience. .
> >>
> >> Thanks,
> >> Harry
> >> --
> >> View this message in context:
> >>
> http://old.nabble.com/Performance-Test-Setup-for-a-web-site...-tp26223743p262
> >> 23743.html
> >> Sent from the JMeter - User mailing list archive at Nabble.com.
> >>
> >>
> >> ---------------------------------------------------------------------
> >> To unsubscribe, e-mail: jmeter-user-unsubscribe@...
> >> For additional commands, e-mail: jmeter-user-help@...
> >>
> >>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@...
> For additional commands, e-mail: jmeter-user-help@...
>
>

[sebb-2-2]

On 06/11/2009, Deepak Shetty <shettyd@...> wrote:

> Hi
>  one of the key problems that I face is simulating varying concurrent usage
>  accurately. e.g. when the day starts out you want this to be low and hit
>  peaks as people come into work and then drop off at lunch or whatever.
>  Specifying different thread groups is one of the ways that you simulate this
>  as you have pointed out.
>  JMeter doesn't have a way to vary thread counts in a thread group easily ,
>  you must setup the total number a priori and then use ramp ups or delays or
>  whatever.
>  Is there are any plans to be able to vary the number of active threads in
>  the threadgroup at runtime?
>  regards

Probably not in the next release, but perhaps after that.

You can use the Constant Throughput Timer in conjunction with a
variable in order to vary the load, see for example:

http://jakarta.apache.org/jmeter/usermanual/best-practices.html#beanshell_server

I've used that in performance tests that ran over many hours.

> deepak
>
>
>  On Fri, Nov 6, 2009 at 6:54 AM, Carl Shaulis <cshaulis@...> wrote:
>
>  > Getting proportional scripts I can think of two approaches:
>  > 1.  Scrap a days worth of log files for all of the requests coming to your
>  > application servers then replay these requests using JMeter
>  > 2.  Estimate what portion of your community does what, then use various
>  > thread groups to emulate this traffic (Thread group one (ie-visit home
>  > page)
>  > 50%, ThreadGroup2 (ie. - Search) 40% etc...
>  >
>  > Number of virtual threads really is dependent on a combination of your load
>  > servers and your target.  For example we have a 16 core machine as a load
>  > generator and a an equivalent machine supporting our application.  Using
>  > 100
>  > concurrent threads and no sleeps the load machine was less than 10% CPU
>  > utilization and the target machine was at 90% CPU utilization.  When then
>  > cached a bunch of the requests and the target machine could respond faster
>  > than the load machine so the Load machine CPU was stressed at the same
>  > thread count, additionally we were almost at our bandwidth limits.  So
>  > Deepak is spot on saying monitor machine resources.
>  >
>  > We have had success using MySQL to inspect data, but we have also kept our
>  > load times brief (5 minutes).  If using a linux box to collect your data
>  > you
>  > can get a quick evaluation of throughput by using wc -l <filename> divided
>  > by the test duration to give you TPS.
>  >
>  > Good luck!
>  >
>  > Carl
>  >
>  >
>  > On 11/6/09 1:13 AM, "Deepak Shetty" <shettyd@...> wrote:
>  >
>  > >> I will create Scripts which will hit the site pages proportional to
>  > there
>  > > usage
>  > > This isnt easy.
>  > >
>  > >> - What should be the configuration of the machine which will simulate
>  > these
>  > >> - If more then one test machine is required please specify there
>  > > configuration?
>  > >> - How many instances of jMeter we need to run for simulating 5000 users?
>  > > My preference is Multiple low end machines running separate JMeter
>  > instances
>  > > to 1 big machine. It simulates the network better . The load you can
>  > > generate depends on what else is running and what your tests actually do.
>  > > People have reported running 1000 threads from a single machine. In any
>  > case
>  > > , you must generate a load and check your client machine resources ,
>  > > preferably the cpu shouldn't exceed 60-80% and memory used should all be
>  > RAM
>  > > not virtual. You can increase the number of threads till you hit some
>  > limit
>  > > after which the client machine may become a bottleneck and give you
>  > > incorrect results.
>  > > See related
>  > > http://wiki.apache.org/jakarta-jmeter/HowManyThreads
>  > >
>  > http://jakarta.apache.org/jmeter/usermanual/best-practices.html#lean_mean
>  > > When running multiple jmeter instances you can either run each instance
>  > > separately (my preference) or you can run Jmeter  in master slave (this
>  > is
>  > > more inefficient , check the mail archives).
>  > >
>  > >> How the result file output should be consolidated from various scripts?
>  > If
>  > >> someone is using any tool for consolidating the output files please
>  > share
>  > >> information regarding the same.
>  > >> - Is there any tool for converting these files to some reports. I have
>  > > heard
>  > >> that reports can be generated using some available xslt's in ant. But
>  > that
>  > >> does not work with large output files. So please suggest alternatives.
>  > > I assume you mean combining the results if you run jmeter instances
>  > > separately. if you use CSV as your format, just concatenate. Mostly if
>  > you
>  > > have long running tests , you would load the CSV files into a database
>  > > table. You have some listeners that can read the CSV/JTL files , but
>  > you'd
>  > > have to use a listener whose memory utlization is constant and not
>  > > proportional  to number of samples (e.g. summary report -
>  > > http://jakarta.apache.org/jmeter/usermanual/component_reference.htmlsection
>  > > 18.3).
>  > >
>  > > regards
>  > > deepak
>  > >
>  > >
>  > > On Thu, Nov 5, 2009 at 2:54 PM, Harry_ <harjitworks@...> wrote:
>  > >
>  > >>
>  > >> Hi,
>  > >>
>  > >> We need to do performance testing for our website simulating 5000 users
>  > >> using jmeter and other open source tools. The following information is
>  > with
>  > >> us:
>  > >>
>  > >> - A csv file containing links and number of times that link last month.
>  > The
>  > >> file is sorted according to popularity of page visited.
>  > >> - With this file I can get information about average number of hits per
>  > >> unit
>  > >> time say per minute. (estimate can be made regarding max load, min
>  > load). I
>  > >> will create Scripts which will hit the site pages proportional to there
>  > >> usage.
>  > >> - All the requests will be http requests.
>  > >> - Average size of the page will be 350 KB (including embedded objects
>  > >> within
>  > >> the page).
>  > >>
>  > >> Now we need answer to the following questions:
>  > >>
>  > >> - How should performance testing of the site simulating that much user
>  > load
>  > >> using Jmeter should be done?
>  > >> - What should be the configuration of the machine which will simulate
>  > these
>  > >> many users? (Someone suggested me 8 core@3 GHz, 16 GB machine).
>  > >> - If more then one test machine is required please specify there
>  > >> configuration?
>  > >> - How many instances of jMeter we need to run for simulating 5000 users?
>  > >> - How the result file output should be consolidated from various
>  > scripts?
>  > >> If
>  > >> someone is using any tool for consolidating the output files please
>  > share
>  > >> information regarding the same.
>  > >> - Is there any tool for converting these files to some reports. I have
>  > >> heard
>  > >> that reports can be generated using some available xslt's in ant. But
>  > that
>  > >> does not work with large output files. So please suggest alternatives.
>  > >> - What other things should i keep in mind for doing the performance
>  > test?
>  > >>
>  > >> We would appreciate if someone can answer these queries based on there
>  > >> experience. .
>  > >>
>  > >> Thanks,
>  > >> Harry
>  > >> --
>  > >> View this message in context:
>  > >>
>  > http://old.nabble.com/Performance-Test-Setup-for-a-web-site...-tp26223743p262
>  > >> 23743.html
>  > >> Sent from the JMeter - User mailing list archive at Nabble.com.
>  > >>
>  > >>
>  > >> ---------------------------------------------------------------------
>  > >> To unsubscribe, e-mail: jmeter-user-unsubscribe@...
>  > >> For additional commands, e-mail: jmeter-user-help@...
>  > >>
>  > >>
>  >
>  >
>  > ---------------------------------------------------------------------
>  > To unsubscribe, e-mail: jmeter-user-unsubscribe@...
>  > For additional commands, e-mail: jmeter-user-help@...
>  >
>  >
>

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

[Deepak Shetty]

ok cool, I did not know you could do that
regards
deepak

On Fri, Nov 6, 2009 at 10:08 AM, sebb <sebbaz@...> wrote:

> On 06/11/2009, Deepak Shetty <shettyd@...> wrote:
> > Hi
> >  one of the key problems that I face is simulating varying concurrent
> usage
> >  accurately. e.g. when the day starts out you want this to be low and hit
> >  peaks as people come into work and then drop off at lunch or whatever.
> >  Specifying different thread groups is one of the ways that you simulate
> this
> >  as you have pointed out.
> >  JMeter doesn't have a way to vary thread counts in a thread group easily
> ,
> >  you must setup the total number a priori and then use ramp ups or delays
> or
> >  whatever.
> >  Is there are any plans to be able to vary the number of active threads
> in
> >  the threadgroup at runtime?
> >  regards
>
> Probably not in the next release, but perhaps after that.
>
> You can use the Constant Throughput Timer in conjunction with a
> variable in order to vary the load, see for example:
>
>
> http://jakarta.apache.org/jmeter/usermanual/best-practices.html#beanshell_server
>
> I've used that in performance tests that ran over many hours.
>
> > deepak
> >
> >
> >  On Fri, Nov 6, 2009 at 6:54 AM, Carl Shaulis <cshaulis@...>
> wrote:
> >
> >  > Getting proportional scripts I can think of two approaches:
> >  > 1.  Scrap a days worth of log files for all of the requests coming to
> your
> >  > application servers then replay these requests using JMeter
> >  > 2.  Estimate what portion of your community does what, then use
> various
> >  > thread groups to emulate this traffic (Thread group one (ie-visit home
> >  > page)
> >  > 50%, ThreadGroup2 (ie. - Search) 40% etc...
> >  >
> >  > Number of virtual threads really is dependent on a combination of your
> load
> >  > servers and your target.  For example we have a 16 core machine as a
> load
> >  > generator and a an equivalent machine supporting our application.
>  Using
> >  > 100
> >  > concurrent threads and no sleeps the load machine was less than 10%
> CPU
> >  > utilization and the target machine was at 90% CPU utilization.  When
> then
> >  > cached a bunch of the requests and the target machine could respond
> faster
> >  > than the load machine so the Load machine CPU was stressed at the same
> >  > thread count, additionally we were almost at our bandwidth limits.  So
> >  > Deepak is spot on saying monitor machine resources.
> >  >
> >  > We have had success using MySQL to inspect data, but we have also kept
> our
> >  > load times brief (5 minutes).  If using a linux box to collect your
> data
> >  > you
> >  > can get a quick evaluation of throughput by using wc -l <filename>
> divided
> >  > by the test duration to give you TPS.
> >  >
> >  > Good luck!
> >  >
> >  > Carl
> >  >
> >  >
> >  > On 11/6/09 1:13 AM, "Deepak Shetty" <shettyd@...> wrote:
> >  >
> >  > >> I will create Scripts which will hit the site pages proportional to
> >  > there
> >  > > usage
> >  > > This isnt easy.
> >  > >
> >  > >> - What should be the configuration of the machine which will
> simulate
> >  > these
> >  > >> - If more then one test machine is required please specify there
> >  > > configuration?
> >  > >> - How many instances of jMeter we need to run for simulating 5000
> users?
> >  > > My preference is Multiple low end machines running separate JMeter
> >  > instances
> >  > > to 1 big machine. It simulates the network better . The load you can
> >  > > generate depends on what else is running and what your tests
> actually do.
> >  > > People have reported running 1000 threads from a single machine. In
> any
> >  > case
> >  > > , you must generate a load and check your client machine resources ,
> >  > > preferably the cpu shouldn't exceed 60-80% and memory used should
> all be
> >  > RAM
> >  > > not virtual. You can increase the number of threads till you hit
> some
> >  > limit
> >  > > after which the client machine may become a bottleneck and give you
> >  > > incorrect results.
> >  > > See related
> >  > > http://wiki.apache.org/jakarta-jmeter/HowManyThreads
> >  > >
> >  >
> http://jakarta.apache.org/jmeter/usermanual/best-practices.html#lean_mean
> >  > > When running multiple jmeter instances you can either run each
> instance
> >  > > separately (my preference) or you can run Jmeter  in master slave
> (this
> >  > is
> >  > > more inefficient , check the mail archives).
> >  > >
> >  > >> How the result file output should be consolidated from various
> scripts?
> >  > If
> >  > >> someone is using any tool for consolidating the output files please
> >  > share
> >  > >> information regarding the same.
> >  > >> - Is there any tool for converting these files to some reports. I
> have
> >  > > heard
> >  > >> that reports can be generated using some available xslt's in ant.
> But
> >  > that
> >  > >> does not work with large output files. So please suggest
> alternatives.
> >  > > I assume you mean combining the results if you run jmeter instances
> >  > > separately. if you use CSV as your format, just concatenate. Mostly
> if
> >  > you
> >  > > have long running tests , you would load the CSV files into a
> database
> >  > > table. You have some listeners that can read the CSV/JTL files , but
> >  > you'd
> >  > > have to use a listener whose memory utlization is constant and not
> >  > > proportional  to number of samples (e.g. summary report -
> >  > >
> http://jakarta.apache.org/jmeter/usermanual/component_reference.htmlsection
> >  > > 18.3).
> >  > >
> >  > > regards
> >  > > deepak
> >  > >
> >  > >
> >  > > On Thu, Nov 5, 2009 at 2:54 PM, Harry_ <harjitworks@...>
> wrote:
> >  > >
> >  > >>
> >  > >> Hi,
> >  > >>
> >  > >> We need to do performance testing for our website simulating 5000
> users
> >  > >> using jmeter and other open source tools. The following information
> is
> >  > with
> >  > >> us:
> >  > >>
> >  > >> - A csv file containing links and number of times that link last
> month.
> >  > The
> >  > >> file is sorted according to popularity of page visited.
> >  > >> - With this file I can get information about average number of hits
> per
> >  > >> unit
> >  > >> time say per minute. (estimate can be made regarding max load, min
> >  > load). I
> >  > >> will create Scripts which will hit the site pages proportional to
> there
> >  > >> usage.
> >  > >> - All the requests will be http requests.
> >  > >> - Average size of the page will be 350 KB (including embedded
> objects
> >  > >> within
> >  > >> the page).
> >  > >>
> >  > >> Now we need answer to the following questions:
> >  > >>
> >  > >> - How should performance testing of the site simulating that much
> user
> >  > load
> >  > >> using Jmeter should be done?
> >  > >> - What should be the configuration of the machine which will
> simulate
> >  > these
> >  > >> many users? (Someone suggested me 8 core@3 GHz, 16 GB machine).
> >  > >> - If more then one test machine is required please specify there
> >  > >> configuration?
> >  > >> - How many instances of jMeter we need to run for simulating 5000
> users?
> >  > >> - How the result file output should be consolidated from various
> >  > scripts?
> >  > >> If
> >  > >> someone is using any tool for consolidating the output files please
> >  > share
> >  > >> information regarding the same.
> >  > >> - Is there any tool for converting these files to some reports. I
> have
> >  > >> heard
> >  > >> that reports can be generated using some available xslt's in ant.
> But
> >  > that
> >  > >> does not work with large output files. So please suggest
> alternatives.
> >  > >> - What other things should i keep in mind for doing the performance
> >  > test?
> >  > >>
> >  > >> We would appreciate if someone can answer these queries based on
> there
> >  > >> experience. .
> >  > >>
> >  > >> Thanks,
> >  > >> Harry
> >  > >> --
> >  > >> View this message in context:
> >  > >>
> >  >
> http://old.nabble.com/Performance-Test-Setup-for-a-web-site...-tp26223743p262
> >  > >> 23743.html
> >  > >> Sent from the JMeter - User mailing list archive at Nabble.com.
> >  > >>
> >  > >>
> >  > >>
> ---------------------------------------------------------------------
> >  > >> To unsubscribe, e-mail: jmeter-user-unsubscribe@...
> >  > >> For additional commands, e-mail:
> jmeter-user-help@...
> >  > >>
> >  > >>
> >  >
> >  >
> >  > ---------------------------------------------------------------------
> >  > To unsubscribe, e-mail: jmeter-user-unsubscribe@...
> >  > For additional commands, e-mail: jmeter-user-help@...
> >  >
> >  >
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@...
> For additional commands, e-mail: jmeter-user-help@...
>
>

[Harry_]

Thank you very much for replying with such a useful information.

Based on all the suggestions, We are creating the scripts which will hit the server proportional to there usage.

Second thing now is to run these scripts through ant. (Will run multiple tests in parallel with each test running for some amount of time in one JVM, then a new JVM instance will be called to further continue the test. We will do this to avoid memory issues which come due to long runs).

and then run schedule the ant scripts through CruiseControl. (As server will be available a night. But at the moment we have no idea how to use it. I have posted seperate threads for questions on ant (http://old.nabble.com/How-to-pass-arguments-to-a-jMeter-jmx-through-ant-and-do-conditional-execution-of-a-target...-td26259404.html) and cruisecontrol (http://old.nabble.com/Scheduling-ant-tasks-through-CruiseControl-td26259822.html). Please reply on the same if you have experience in the same.)

Also do give your recommendations on the approach we are following.

Once again thanks a ton for your priceless replies....

Thanks and Regards,
Harry__

Deepak Shetty wrote:

ok cool, I did not know you could do that
regards
deepak

On Fri, Nov 6, 2009 at 10:08 AM, sebb <[e-mail]> wrote:

> On 06/11/2009, Deepak Shetty <[e-mail]> wrote:
> > Hi
> >  one of the key problems that I face is simulating varying concurrent
> usage
> >  accurately. e.g. when the day starts out you want this to be low and hit
> >  peaks as people come into work and then drop off at lunch or whatever.
> >  Specifying different thread groups is one of the ways that you simulate
> this
> >  as you have pointed out.
> >  JMeter doesn't have a way to vary thread counts in a thread group easily
> ,
> >  you must setup the total number a priori and then use ramp ups or delays
> or
> >  whatever.
> >  Is there are any plans to be able to vary the number of active threads
> in
> >  the threadgroup at runtime?
> >  regards
>
> Probably not in the next release, but perhaps after that.
>
> You can use the Constant Throughput Timer in conjunction with a
> variable in order to vary the load, see for example:
>
>
> http://jakarta.apache.org/jmeter/usermanual/best-practices.html#beanshell_server
>
> I've used that in performance tests that ran over many hours.
>
> > deepak
> >
> >
> >  On Fri, Nov 6, 2009 at 6:54 AM, Carl Shaulis <[e-mail]>
> wrote:
> >
> >  > Getting proportional scripts I can think of two approaches:
> >  > 1.  Scrap a days worth of log files for all of the requests coming to
> your
> >  > application servers then replay these requests using JMeter
> >  > 2.  Estimate what portion of your community does what, then use
> various
> >  > thread groups to emulate this traffic (Thread group one (ie-visit home
> >  > page)
> >  > 50%, ThreadGroup2 (ie. - Search) 40% etc...
> >  >
> >  > Number of virtual threads really is dependent on a combination of your
> load
> >  > servers and your target.  For example we have a 16 core machine as a
> load
> >  > generator and a an equivalent machine supporting our application.
>  Using
> >  > 100
> >  > concurrent threads and no sleeps the load machine was less than 10%
> CPU
> >  > utilization and the target machine was at 90% CPU utilization.  When
> then
> >  > cached a bunch of the requests and the target machine could respond
> faster
> >  > than the load machine so the Load machine CPU was stressed at the same
> >  > thread count, additionally we were almost at our bandwidth limits.  So
> >  > Deepak is spot on saying monitor machine resources.
> >  >
> >  > We have had success using MySQL to inspect data, but we have also kept
> our
> >  > load times brief (5 minutes).  If using a linux box to collect your
> data
> >  > you
> >  > can get a quick evaluation of throughput by using wc -l <filename>
> divided
> >  > by the test duration to give you TPS.
> >  >
> >  > Good luck!
> >  >
> >  > Carl
> >  >
> >  >
> >  > On 11/6/09 1:13 AM, "Deepak Shetty" <[e-mail]> wrote:
> >  >
> >  > >> I will create Scripts which will hit the site pages proportional to
> >  > there
> >  > > usage
> >  > > This isnt easy.
> >  > >
> >  > >> - What should be the configuration of the machine which will
> simulate
> >  > these
> >  > >> - If more then one test machine is required please specify there
> >  > > configuration?
> >  > >> - How many instances of jMeter we need to run for simulating 5000
> users?
> >  > > My preference is Multiple low end machines running separate JMeter
> >  > instances
> >  > > to 1 big machine. It simulates the network better . The load you can
> >  > > generate depends on what else is running and what your tests
> actually do.
> >  > > People have reported running 1000 threads from a single machine. In
> any
> >  > case
> >  > > , you must generate a load and check your client machine resources ,
> >  > > preferably the cpu shouldn't exceed 60-80% and memory used should
> all be
> >  > RAM
> >  > > not virtual. You can increase the number of threads till you hit
> some
> >  > limit
> >  > > after which the client machine may become a bottleneck and give you
> >  > > incorrect results.
> >  > > See related
> >  > > http://wiki.apache.org/jakarta-jmeter/HowManyThreads
> >  > >
> >  >
> http://jakarta.apache.org/jmeter/usermanual/best-practices.html#lean_mean
> >  > > When running multiple jmeter instances you can either run each
> instance
> >  > > separately (my preference) or you can run Jmeter  in master slave
> (this
> >  > is
> >  > > more inefficient , check the mail archives).
> >  > >
> >  > >> How the result file output should be consolidated from various
> scripts?
> >  > If
> >  > >> someone is using any tool for consolidating the output files please
> >  > share
> >  > >> information regarding the same.
> >  > >> - Is there any tool for converting these files to some reports. I
> have
> >  > > heard
> >  > >> that reports can be generated using some available xslt's in ant.
> But
> >  > that
> >  > >> does not work with large output files. So please suggest
> alternatives.
> >  > > I assume you mean combining the results if you run jmeter instances
> >  > > separately. if you use CSV as your format, just concatenate. Mostly
> if
> >  > you
> >  > > have long running tests , you would load the CSV files into a
> database
> >  > > table. You have some listeners that can read the CSV/JTL files , but
> >  > you'd
> >  > > have to use a listener whose memory utlization is constant and not
> >  > > proportional  to number of samples (e.g. summary report -
> >  > >
> http://jakarta.apache.org/jmeter/usermanual/component_reference.htmlsection
> >  > > 18.3).
> >  > >
> >  > > regards
> >  > > deepak
> >  > >
> >  > >
> >  > > On Thu, Nov 5, 2009 at 2:54 PM, Harry_ <[e-mail]>
> wrote:
> >  > >
> >  > >>
> >  > >> Hi,
> >  > >>
> >  > >> We need to do performance testing for our website simulating 5000
> users
> >  > >> using jmeter and other open source tools. The following information
> is
> >  > with
> >  > >> us:
> >  > >>
> >  > >> - A csv file containing links and number of times that link last
> month.
> >  > The
> >  > >> file is sorted according to popularity of page visited.
> >  > >> - With this file I can get information about average number of hits
> per
> >  > >> unit
> >  > >> time say per minute. (estimate can be made regarding max load, min
> >  > load). I
> >  > >> will create Scripts which will hit the site pages proportional to
> there
> >  > >> usage.
> >  > >> - All the requests will be http requests.
> >  > >> - Average size of the page will be 350 KB (including embedded
> objects
> >  > >> within
> >  > >> the page).
> >  > >>
> >  > >> Now we need answer to the following questions:
> >  > >>
> >  > >> - How should performance testing of the site simulating that much
> user
> >  > load
> >  > >> using Jmeter should be done?
> >  > >> - What should be the configuration of the machine which will
> simulate
> >  > these
> >  > >> many users? (Someone suggested me 8 core@3 GHz, 16 GB machine).
> >  > >> - If more then one test machine is required please specify there
> >  > >> configuration?
> >  > >> - How many instances of jMeter we need to run for simulating 5000
> users?
> >  > >> - How the result file output should be consolidated from various
> >  > scripts?
> >  > >> If
> >  > >> someone is using any tool for consolidating the output files please
> >  > share
> >  > >> information regarding the same.
> >  > >> - Is there any tool for converting these files to some reports. I
> have
> >  > >> heard
> >  > >> that reports can be generated using some available xslt's in ant.
> But
> >  > that
> >  > >> does not work with large output files. So please suggest
> alternatives.
> >  > >> - What other things should i keep in mind for doing the performance
> >  > test?
> >  > >>
> >  > >> We would appreciate if someone can answer these queries based on
> there
> >  > >> experience. .
> >  > >>
> >  > >> Thanks,
> >  > >> Harry
> >  > >> --
> >  > >> View this message in context:
> >  > >>
> >  >
> http://old.nabble.com/Performance-Test-Setup-for-a-web-site...-tp26223743p262
> >  > >> 23743.html
> >  > >> Sent from the JMeter - User mailing list archive at Nabble.com.
> >  > >>
> >  > >>
> >  > >>
> ---------------------------------------------------------------------
> >  > >> To unsubscribe, e-mail: [e-mail]
> >  > >> For additional commands, e-mail:
> [e-mail]
> >  > >>
> >  > >>
> >  >
> >  >
> >  > ---------------------------------------------------------------------
> >  > To unsubscribe, e-mail: [e-mail]
> >  > For additional commands, e-mail: [e-mail]
> >  >
> >  >
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [e-mail]
> For additional commands, e-mail: [e-mail]
>
>

[André van Hoorn-2]

Hi,

> On 06/11/2009, Deepak Shetty <shettyd@...> wrote:
>> Hi
>>  one of the key problems that I face is simulating varying concurrent usage
>>  accurately. e.g. when the day starts out you want this to be low and hit
>>  peaks as people come into work and then drop off at lunch or whatever.
>>  Specifying different thread groups is one of the ways that you simulate this
>>  as you have pointed out.
>>  JMeter doesn't have a way to vary thread counts in a thread group easily ,
>>  you must setup the total number a priori and then use ramp ups or delays or
>>  whatever.
>>  Is there are any plans to be able to vary the number of active threads in
>>  the threadgroup at runtime?
>>  regards
>
> Probably not in the next release, but perhaps after that.
>
> You can use the Constant Throughput Timer in conjunction with a
> variable in order to vary the load, see for example:
>
> http://jakarta.apache.org/jmeter/usermanual/best-practices.html#beanshell_server
>
> I've used that in performance tests that ran over many hours.

maybe you also want to have a look at our JMeter plug-in Markov4JMeter
(http://markov4jmeter.sourceforge.net/) which allows to vary the
workload intensity (in terms of active users/threads) based on
mathematic formaluae (implemented based on JMeter's BeanShell
integration). We are using this in our research experiments to simulate
varying workloads.

Best regards,
André

>
>> deepak
>>
>>
>>  On Fri, Nov 6, 2009 at 6:54 AM, Carl Shaulis <cshaulis@...> wrote:
>>
>>  > Getting proportional scripts I can think of two approaches:
>>  > 1.  Scrap a days worth of log files for all of the requests coming to your
>>  > application servers then replay these requests using JMeter
>>  > 2.  Estimate what portion of your community does what, then use various
>>  > thread groups to emulate this traffic (Thread group one (ie-visit home
>>  > page)
>>  > 50%, ThreadGroup2 (ie. - Search) 40% etc...
>>  >
>>  > Number of virtual threads really is dependent on a combination of your load
>>  > servers and your target.  For example we have a 16 core machine as a load
>>  > generator and a an equivalent machine supporting our application.  Using
>>  > 100
>>  > concurrent threads and no sleeps the load machine was less than 10% CPU
>>  > utilization and the target machine was at 90% CPU utilization.  When then
>>  > cached a bunch of the requests and the target machine could respond faster
>>  > than the load machine so the Load machine CPU was stressed at the same
>>  > thread count, additionally we were almost at our bandwidth limits.  So
>>  > Deepak is spot on saying monitor machine resources.
>>  >
>>  > We have had success using MySQL to inspect data, but we have also kept our
>>  > load times brief (5 minutes).  If using a linux box to collect your data
>>  > you
>>  > can get a quick evaluation of throughput by using wc -l <filename> divided
>>  > by the test duration to give you TPS.
>>  >
>>  > Good luck!
>>  >
>>  > Carl
>>  >
>>  >
>>  > On 11/6/09 1:13 AM, "Deepak Shetty" <shettyd@...> wrote:
>>  >
>>  > >> I will create Scripts which will hit the site pages proportional to
>>  > there
>>  > > usage
>>  > > This isnt easy.
>>  > >
>>  > >> - What should be the configuration of the machine which will simulate
>>  > these
>>  > >> - If more then one test machine is required please specify there
>>  > > configuration?
>>  > >> - How many instances of jMeter we need to run for simulating 5000 users?
>>  > > My preference is Multiple low end machines running separate JMeter
>>  > instances
>>  > > to 1 big machine. It simulates the network better . The load you can
>>  > > generate depends on what else is running and what your tests actually do.
>>  > > People have reported running 1000 threads from a single machine. In any
>>  > case
>>  > > , you must generate a load and check your client machine resources ,
>>  > > preferably the cpu shouldn't exceed 60-80% and memory used should all be
>>  > RAM
>>  > > not virtual. You can increase the number of threads till you hit some
>>  > limit
>>  > > after which the client machine may become a bottleneck and give you
>>  > > incorrect results.
>>  > > See related
>>  > > http://wiki.apache.org/jakarta-jmeter/HowManyThreads
>>  > >
>>  > http://jakarta.apache.org/jmeter/usermanual/best-practices.html#lean_mean
>>  > > When running multiple jmeter instances you can either run each instance
>>  > > separately (my preference) or you can run Jmeter  in master slave (this
>>  > is
>>  > > more inefficient , check the mail archives).
>>  > >
>>  > >> How the result file output should be consolidated from various scripts?
>>  > If
>>  > >> someone is using any tool for consolidating the output files please
>>  > share
>>  > >> information regarding the same.
>>  > >> - Is there any tool for converting these files to some reports. I have
>>  > > heard
>>  > >> that reports can be generated using some available xslt's in ant. But
>>  > that
>>  > >> does not work with large output files. So please suggest alternatives.
>>  > > I assume you mean combining the results if you run jmeter instances
>>  > > separately. if you use CSV as your format, just concatenate. Mostly if
>>  > you
>>  > > have long running tests , you would load the CSV files into a database
>>  > > table. You have some listeners that can read the CSV/JTL files , but
>>  > you'd
>>  > > have to use a listener whose memory utlization is constant and not
>>  > > proportional  to number of samples (e.g. summary report -
>>  > > http://jakarta.apache.org/jmeter/usermanual/component_reference.htmlsection
>>  > > 18.3).
>>  > >
>>  > > regards
>>  > > deepak
>>  > >
>>  > >
>>  > > On Thu, Nov 5, 2009 at 2:54 PM, Harry_ <harjitworks@...> wrote:
>>  > >
>>  > >>
>>  > >> Hi,
>>  > >>
>>  > >> We need to do performance testing for our website simulating 5000 users
>>  > >> using jmeter and other open source tools. The following information is
>>  > with
>>  > >> us:
>>  > >>
>>  > >> - A csv file containing links and number of times that link last month.
>>  > The
>>  > >> file is sorted according to popularity of page visited.
>>  > >> - With this file I can get information about average number of hits per
>>  > >> unit
>>  > >> time say per minute. (estimate can be made regarding max load, min
>>  > load). I
>>  > >> will create Scripts which will hit the site pages proportional to there
>>  > >> usage.
>>  > >> - All the requests will be http requests.
>>  > >> - Average size of the page will be 350 KB (including embedded objects
>>  > >> within
>>  > >> the page).
>>  > >>
>>  > >> Now we need answer to the following questions:
>>  > >>
>>  > >> - How should performance testing of the site simulating that much user
>>  > load
>>  > >> using Jmeter should be done?
>>  > >> - What should be the configuration of the machine which will simulate
>>  > these
>>  > >> many users? (Someone suggested me 8 core@3 GHz, 16 GB machine).
>>  > >> - If more then one test machine is required please specify there
>>  > >> configuration?
>>  > >> - How many instances of jMeter we need to run for simulating 5000 users?
>>  > >> - How the result file output should be consolidated from various
>>  > scripts?
>>  > >> If
>>  > >> someone is using any tool for consolidating the output files please
>>  > share
>>  > >> information regarding the same.
>>  > >> - Is there any tool for converting these files to some reports. I have
>>  > >> heard
>>  > >> that reports can be generated using some available xslt's in ant. But
>>  > that
>>  > >> does not work with large output files. So please suggest alternatives.
>>  > >> - What other things should i keep in mind for doing the performance
>>  > test?
>>  > >>
>>  > >> We would appreciate if someone can answer these queries based on there
>>  > >> experience. .
>>  > >>
>>  > >> Thanks,
>>  > >> Harry
>>  > >> --
>>  > >> View this message in context:
>>  > >>
>>  > http://old.nabble.com/Performance-Test-Setup-for-a-web-site...-tp26223743p262
>>  > >> 23743.html
>>  > >> Sent from the JMeter - User mailing list archive at Nabble.com.
>>  > >>
>>  > >>
>>  > >> ---------------------------------------------------------------------
>>  > >> To unsubscribe, e-mail: jmeter-user-unsubscribe@...
>>  > >> For additional commands, e-mail: jmeter-user-help@...
>>  > >>
>>  > >>
>>  >
>>  >
>>  > ---------------------------------------------------------------------
>>  > To unsubscribe, e-mail: jmeter-user-unsubscribe@...
>>  > For additional commands, e-mail: jmeter-user-help@...
>>  >
>>  >
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@...
> For additional commands, e-mail: jmeter-user-help@...
>

--
Dipl.-Inform. André van Hoorn

DFG Graduate School on Trustworthy Software Systems (TrustSoft)
Department of Computer Science, University of Oldenburg
PO Box 2503, D-26111 Oldenburg, Germany

Room A2 2-224, Tel: [phone], Fax: -2196
WWW:     http://www.trustsoft.uni-oldenburg.de/members/andrevh
E-Mail:  Andre.van.Hoorn@...
PGP key: http://www.trustsoft.uni-oldenburg.de/members/andrevh#pgp

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