How Frustration Co-efficient will calculated?
zameer_india_ants
Posts: 5
Hi... All...
I tested successfully for our webiste... but i found a message like
Ur Co's webiste Frustration Coefficient was 50%.
Now I have one doubt...
On what factors/metrics this Frustration Coefficient is calculated?
I mean from the result file we have summary informations , in which we have some common factors/metrics like "Time to Connect", "Time to first byte" and "Time to last byte".
Now My Question was
1.) Whether these factors/metrics involve to calculate Frustration
Coefficient. or not?
Second Question :
2.) What is the max time of "Time to last byte".
3.) Can we expect if a last byte reaches more than the maximum time leads to Frustration Coefficient?
Actually I would like to anayze the result file.. when aim trying to analyze the file , i am not getting more info rather than pages time to connect, similarlly for objects.performance counters etc..
but if the result file gives much more help like In which areas (code,bulkdata,server architecture,what to upgrade depend on application)where our application is weak. and what to do in increase performance?
and one more question ...
4.) This performance Encounter shows whether Our sytem's graph or Server Sytem Graph?
Any Help Would be appreciate..
I tested successfully for our webiste... but i found a message like
Ur Co's webiste Frustration Coefficient was 50%.
Now I have one doubt...
On what factors/metrics this Frustration Coefficient is calculated?
I mean from the result file we have summary informations , in which we have some common factors/metrics like "Time to Connect", "Time to first byte" and "Time to last byte".
Now My Question was
1.) Whether these factors/metrics involve to calculate Frustration
Coefficient. or not?
Second Question :
2.) What is the max time of "Time to last byte".
3.) Can we expect if a last byte reaches more than the maximum time leads to Frustration Coefficient?
Actually I would like to anayze the result file.. when aim trying to analyze the file , i am not getting more info rather than pages time to connect, similarlly for objects.performance counters etc..
but if the result file gives much more help like In which areas (code,bulkdata,server architecture,what to upgrade depend on application)where our application is weak. and what to do in increase performance?
and one more question ...
4.) This performance Encounter shows whether Our sytem's graph or Server Sytem Graph?
Any Help Would be appreciate..
Thanx in Advance,
zameer ahmed syed.
zameer ahmed syed.
This discussion has been closed.
Comments
There was a bit of research involved in designing the frustration coefficient and abandonment predictions. For more information, you may want to have a look at
http://www.simple-talk.com/2003/03/13/t ... rformance/
This article describes the whole process used to come up with the frustration coefficient.
It is calculated on the time to first and last byte. Mainly, your website user needs to see some information in a certain number of seconds.
ANTS Load also lets you set a maximum time to last byte in the edit load dialogue. When 95% of the requests fall out of this metric, the test will be marked as failed in the report. This is separate from the frustration coefficient, which is based on the statistics that we ourselves have collected. This is because in some cases, you may decide that longer wait times are more acceptable.
Perhaps in the future, it may be a good idea for the ANTS Reports to point out that, for instance, the processor was not working too hard, but there were long disk queues, please consider upgrading your hard disk, or something similar to that. ANTS Load will not do those things for you at the moment.
It's much better at helping you find the 'slow pages' so that perhaps you can optimize the code, or put more information into the session rather than making trips to the database back-end as often.
U gave me good and theoritical and descriptive explanation. Thanq very much.
---zameer ahmed syed.
zameer ahmed syed.