Unanswered question

Neoload graphs shows higher avg response time than values section

I ran a load test and while analyzing the results I found out that the Transaction/Values section of my summary had response time values lesser than that of my graphs.
I prepared Avg Duration custom graphs for my transaction. when i saw the value in the legend it was a bit higher than what i have in values/transaction section.

Eg:
Values section: Trans 1: Avg repsonse time :26.5 sec
Graph section: Trans 1: Avg Duration :27 sec

Values section: Trans 2: Avg repsonse time :19.9 sec
Graph section: Trans 1: Avg Duration :20.183 sec

Can anyone help me here?

Rishabh W.
Rishabh W.

Rishabh W.

Level
0
34 / 100
points

Answers

Nouredine A.
Nouredine A.

Nouredine A.

Level
4
5000 / 5000
points
Team

The average from the Values tab is the accurate one. It is computed from the raw values. You can compute it again by exporting the raw data of your transaction from the Values tab to a CSV file.

Regarding the average of the graph section it's the average from the values in that graph. Since on the graph you do not have the raw values but data aggregated by time interval the average on the graph is different due to rounded values.

Rishabh W.
Rishabh W.

Rishabh W.

Level
0
34 / 100
points

Thanks Nouredine.

I had another ques, if i have to test the system with 1000 user load and each performing 2 transactions, can that be achieved with 250 user license? my client has 250 VU license but needs 1000 concurrent users on the system

Nouredine A.
Nouredine A.

Nouredine A.

Level
4
5000 / 5000
points
Team

A 250 VU license allows you to run 250 virtual users concurrently and not more. You need a 1000VU license to run 1000 concurrent users.

Rishabh W.
Rishabh W.

Rishabh W.

Level
0
34 / 100
points

Thanks,

One more thing. i have a page which has 50-80 correlations, most of the values have same boundaries but values are different at different occurrence.
right now I am manually correlating them but is there a way to use regular expression to correlate them? also how do i replace particular occurrence with value which is expected at that place.

all the values are found on 1 page

Nouredine A.
Nouredine A.

Nouredine A.

Level
4
5000 / 5000
points
Team

You should be able to manage that with regular expression even if the boundaries are not the same all the time. But without seeing the response and the values you would like to extract it's difficult to provide you more details.

In any case the recommanded way is to always use framework parameter when possible. First you won't have to do such work all the time and the framework will automatically replace the occurrences for you. In the framework definition you can specify where exactly the values should be replaced. For that you can use regexp or XPath and in our next version coming this week JSON path.

 

Rishabh W.
Rishabh W.

Rishabh W.

Level
0
34 / 100
points

I cant paste the actual response here but below is the sample.
the entity id occurs 50 times in response and its boundaries are same always. since the occurrence is different hence value is different and hence i need to manually correlate it to replace a particular value in next page. all the 50 occurrences are on 1 page.

\"EntityId\":\"d52aec44-f771-4cc8-aa8c-d9f7e8d90717\",

,\"EntityId\":\"0cb52fa4-d58b-4f6b-9e6d-6803f003bf7e\",

\"EntityId\":\"b8488097-39dd-41dd-bb10-59f5e3b56f0f\",

\"EntityId\":\"d207ab6b-d2b1-4e68-bff9-96f41bacbe24\",

Nouredine A.
Nouredine A.

Nouredine A.

Level
4
5000 / 5000
points
Team

With a variable extractor configured to extract all those EntityId values and if you moved it as framework parameter NeoLoad will automatically replaced all the occurrences in your requests in the same order.

That means that it will do exactly what you do but automatically.