RE: [analog-help] Process Time and request reports

[analog-help]

Title: Process Time and request reports

'Lo Jo,

 

  I'm not at all certain if PROCTIME includes the wait that a request might endure before being serviced.  So you may still be

somewhat shy of your goal.  If it's pure server benchmarking you're after, ANALOG may still fall short

 

  But don't get me wrong.. RE: WebLogAnalysis... nothing beats ANALOG's power/performance.

 

 Cheers!

 

 - - -

 Micke

-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Josephine Dyer
Sent: Monday, September 30, 2002 7:03 PM
To: '[EMAIL PROTECTED]'
Subject: RE: [analog-help] Process Time and request reports

Hi Micke,

 

Thanks for your help.  Apologies though as I should have been a little clearer.  We are concerned about the performance of IIS and the application server over periods of time, and the effects of content deployments, code changes, and application server tweaks, so we are really interested in measuring how long it takes IIS to service the request in order that we can perform comparisons between days / hours.  We also need to be able to compare the request servicing times of our site in both production and contingency situations.

 

Thanks again for pointing me to the other tool though as it is always good to look at recommended tools.

 

Thanks,

 

Jo 

 

From: Micke Barnes [mailto:[EMAIL PROTECTED]]
Sent: Monday, 30 September 2002 9:35 PM
To: [EMAIL PROTECTED]
Subject: RE: [analog-help] Process Time and request reports

    'Lo Jo,

 

    PROCESSING TIME is not a measurement of how long it takes for the web page

    to download to the client.  PROCTIME is a measurement of how long it took

    IIS to service the request.

 

    If it is your goal to measure the webserver performance from a browser's point

    of view, --Such as Time To Last Byte (TTLB),  then I recommend a different tool.

    Microsoft Web Application Stress Tool is available at:

    http://webtool.rte.microsoft.com/ and like Analog, is completely free.

 

    Though, in consideration of this list, further discussion of should be continued

    offline.

 

    Sincerely,

 

    Micke Barnes

    [EMAIL PROTECTED]

 

   

 

 

       

-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Josephine Dyer
Sent: Monday, September 30, 2002 1:18 AM
To: '[EMAIL PROTECTED]'
Subject: [analog-help] Process Time and request reports

Hi,

I have been asked by my company to provide a tool that will enable us to track the performance of our website daily using the data in our log files. Page download time is of particular issue and we would like to be able to make comparisons between different days / hours. I have recently downloaded version 5.24 of analog, which I believe to be an exceptionally fast and flexible tool.  I particularly like the ability to call the program via the command line and pass in configuration that is only relevant to that particular call.  (I realise I could also do this via seperate config files)

We are using IIS 5.0.  I have configured the log files to output according to the documentation provided in Configure IIS for use with Analog. I am able to display the sum / percentage of pages it takes to load in a certain amount of time by using:

PROCTIMECOLS P
PROCTIME ON
I am also able to see the most popular pages using the request report. I can report on the dailiy or hourly logs by configuring IIS or (I believe) by setting the time setting in analog.  Need to read some more here.

However this only allows a certain amount of evaluation.  I therefore wrote a batch script that calls analogue via the command line and sets 'FILEINCLUDE' via the command line to allow me to see the processing time of either the most requested files or different directories.  This allows me to compare the average download time for pages in each section / or the most popular pages.  However this does not allow me to see  which pages were particularly slow during that period. I understand that Analog will not cross reference however I am unable to retreive the required information that would allow me to cross reference.

It would be useful to be able to report on the slowest pages to be loaded.  If the request report also returned the time-taken field, I assume I could request a floor of one for this report (REQFLOOR 1r) and then REQSORTBY time taken (this does not appear to be an option)

reqs: %bytes:       last time: file
----: ------: ---------------: ----
   2: 50.82%: 29/Sep/02 23:53: </plan_your_trip/special_travel_needs/singles/special_info_std.aust>
   1:  0.42%: 29/Sep/02 23:52: </plan_your_trip/special_travel_needs/singles/compute_url.aust>
   1:  0.43%: 29/Sep/02 23:49: </plan_your_trip/special_travel_needs/compute_url.aust>
   1: 23.03%: 29/Sep/02 23:49: </plan_your_trip/special_travel_needs/gay_and_lesbian/special_info_std.aust>
   1: 25.30%: 29/Sep/02 23:38: </plan_your_trip/special_travel_needs/singles/special_info_jp_asia.aust>

I have searched through the documentation and mailing list archives for a solution and cannot find anything to help me. 

Thanks in advance for your help.

Jo

Jo Dyer
Java Developer
Australian Tourist Commission
Email: jdyer@atc.australia.com
Phone: +61 2 9361 1224  Fax: +61 2 9361 1270
Corporate Site: www.atc.australia.com
Consumer Site: www.australia.com  



******************************************************************************
The information in this e-mail and any attachments is confidential
and is intended only for the use of the above named recipient.
In certain cases it is also legally privileged. If you are not the
intended recipient, any dissemination, distribution or copying of this
e-mail is strictly prohibited. If you have received this e-mail
in error, please advise immediately by return e-mail.
Unless the contrary is stated, the contents of this message do not
necessarily represent the views or position of the Australian Tourist
Commission. The Australian Tourist Commission does not
represent or warrant that this message or any attachments to it are
free from viruses or defects.
******************************************************************************