Making sense of the test results obtained by running Measure-PnPResponseTime

Regular Contributor



I have been running some performance tests against a pre-production SharePoint 2016 farm.  Anyway, I connect to intranet via (lb) url and execute the  PS snippet below, remotely on one of the dev servers - my supplied laptop is very locked down :(


$testUrls = @{

"IntranetHome" = "/"

"SomeContentRollup" = "/sites/search/pages/myniceresults.aspx?k=sw"

"Pages" = "/Pages/Forms/AllItems.aspx"



$datetimeSuffix = (get-date).ToString("_ddMMyyy_HHmmss")


$count = 1000

$timeout = 50

$warmUp = 10

$histogram = 20


# iterate through key and value pairs

$testUrls.Keys | ForEach-Object{

$_ , $testUrls[$_]

$name = $_

Measure-PnPResponseTime $testUrls[$_] -Count $count -WarmUp $warmUp -Histogram $histogram -Timeout $timeout `

| Select -expa Histogram | % {$_.GetEnumerator() | `

Export-Csv ("C:\Users\sp2013_admin\Documents\Stats\" + $name + "-responsetime-" + $datetimeSuffix + ".csv") -NoTypeInformation}




A colleague and I have been trying to make sense of the histogram data: we know the value adds up to the total number of tests i.e 1000. It's the Key we find confusing - is it some sort of auto scale between min and max values in ms. This might explain why so many keys have zero value.


Also what would a good and a bad set results look like; just so that I can compare. By all means suggest better test urls etc.


The above tests won't run against SharePoint 2007, so I wondered if anyone has used any open source testing tools




4 Replies

@Erwin van Hunen , good talking to you earlier after Vesa's presentation. Here is the question I posted - it is the (key) scale I  find confusing.  


Happy to paste up he CSV file data when I am able to vpn to the office.




Any update on this please ? All we are looking for is an explaination of how the Keys are derived?
best response confirmed by Daniel Westerdale (Regular Contributor)

Hi @Daniel Westerdale,

Take a look at this post in StackOverFlow,, I created and found the solution almost at the same time.



Thanks for the update. This is going back a bit, so it took a while to work out what my 'former self' was asking :smile:.   I think I was tuning a on premises  SP installation at the time . Nevertheless the explanation does make sense.