You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@jmeter.apache.org by Alexander Podelko <ap...@yahoo.com.INVALID> on 2018/09/14 02:17:02 UTC
Monitoring results
Hi,
I am trying to add server monitoring results and got quite a few questions. Sorry for such simple questions - but all numerous posts I found describe how to setup monitoring, but I have found none describing it deeper....
So I setup the agent / jp@gc - PerfMon Metrics Collector and it is working.
By the way, what jp@gc, used in numerous places, stands for?
The first question: how should we configure the files for keeping test results and monitoring results? May they be recorded in the same jtl file or it should be different files?
It appears that for monitoring results the file should be explicitly named inside jp@gc - PerfMon Metrics Collector UI file name filed. Any other way? Doesn't look like we have command line option for that....
Monitoring charts looks nice - any way to see aggregate numbers? Say, average CPU utilization?
Any idea if monitoring data may be fed into Jenkins performance plug-in? Any other idea how to compare resource utilization for two different runs?
Will much appreciate any information on any of above questions.
Thanks,Alex
Re: Monitoring results
Posted by Deepak Shetty <sh...@gmail.com>.
Oh and i believe JP@GC ==> Jmeter plugins @ Google code
regards
deepak
On Mon, Sep 17, 2018 at 1:31 PM Deepak Shetty <sh...@gmail.com> wrote:
> Hi
> Since you are using Jmeter plugins , Im guessing you will find more
> answers on the plugins on their support forum rather than on the JMeter
> mailing list.
>
> JMeter listeners usually have a write results to file , it needs to be its
> own file if I remember correctly. i believe a common pattern is to add
> timestamp to file name or pass in a run number/name. You can use any JMeter
> property for the file name and so you can even set it from command line
> (i.e. set the property and use the property in filename field
> using${__P(name)} function)
>
> If the plugin doesnt do aggregate then you only option is to load the
> result file into something that does or put in a feature request for the
> plugin.
>
> > Any other idea how to compare resource utilization for two different
> runs?
> I believe there is a merge results plugin (there was a thread a week or so
> ago). I usually do these kind of things outside JMeter
>
> Note I havent used this plugin or the Jenkins plugin so I cant really say.
> Usually for the application, one needs to monitor much more than the raw
> OS numbers (like a java VM) and the tools available server side are much
> more sophisticated and provide all the OS numbers correlated to the actual
> running code (YMMV).
>
> regards
> deepak
>
> regards
> deepak
>
> On Thu, Sep 13, 2018 at 7:17 PM Alexander Podelko
> <ap...@yahoo.com.invalid> wrote:
>
>> Hi,
>> I am trying to add server monitoring results and got quite a few
>> questions. Sorry for such simple questions - but all numerous posts I found
>> describe how to setup monitoring, but I have found none describing it
>> deeper....
>> So I setup the agent / jp@gc - PerfMon Metrics Collector and it is
>> working.
>> By the way, what jp@gc, used in numerous places, stands for?
>> The first question: how should we configure the files for keeping test
>> results and monitoring results? May they be recorded in the same jtl file
>> or it should be different files?
>>
>> It appears that for monitoring results the file should be explicitly
>> named inside jp@gc - PerfMon Metrics Collector UI file name filed. Any
>> other way? Doesn't look like we have command line option for that....
>> Monitoring charts looks nice - any way to see aggregate numbers? Say,
>> average CPU utilization?
>> Any idea if monitoring data may be fed into Jenkins performance plug-in?
>> Any other idea how to compare resource utilization for two different runs?
>> Will much appreciate any information on any of above questions.
>> Thanks,Alex
>>
>
Re: Monitoring results
Posted by Deepak Shetty <sh...@gmail.com>.
Hi
Since you are using Jmeter plugins , Im guessing you will find more answers
on the plugins on their support forum rather than on the JMeter mailing
list.
JMeter listeners usually have a write results to file , it needs to be its
own file if I remember correctly. i believe a common pattern is to add
timestamp to file name or pass in a run number/name. You can use any JMeter
property for the file name and so you can even set it from command line
(i.e. set the property and use the property in filename field
using${__P(name)} function)
If the plugin doesnt do aggregate then you only option is to load the
result file into something that does or put in a feature request for the
plugin.
> Any other idea how to compare resource utilization for two different runs?
I believe there is a merge results plugin (there was a thread a week or so
ago). I usually do these kind of things outside JMeter
Note I havent used this plugin or the Jenkins plugin so I cant really say.
Usually for the application, one needs to monitor much more than the raw OS
numbers (like a java VM) and the tools available server side are much more
sophisticated and provide all the OS numbers correlated to the actual
running code (YMMV).
regards
deepak
regards
deepak
On Thu, Sep 13, 2018 at 7:17 PM Alexander Podelko
<ap...@yahoo.com.invalid> wrote:
> Hi,
> I am trying to add server monitoring results and got quite a few
> questions. Sorry for such simple questions - but all numerous posts I found
> describe how to setup monitoring, but I have found none describing it
> deeper....
> So I setup the agent / jp@gc - PerfMon Metrics Collector and it is
> working.
> By the way, what jp@gc, used in numerous places, stands for?
> The first question: how should we configure the files for keeping test
> results and monitoring results? May they be recorded in the same jtl file
> or it should be different files?
>
> It appears that for monitoring results the file should be explicitly named
> inside jp@gc - PerfMon Metrics Collector UI file name filed. Any other
> way? Doesn't look like we have command line option for that....
> Monitoring charts looks nice - any way to see aggregate numbers? Say,
> average CPU utilization?
> Any idea if monitoring data may be fed into Jenkins performance plug-in?
> Any other idea how to compare resource utilization for two different runs?
> Will much appreciate any information on any of above questions.
> Thanks,Alex
>