Charting - purepaths by count and purepaths by top 10 slowest (agg) purepaths
Been using DT for years off and on... I'm looking to create a couple of charts to do the following:
- aggregate and count unique purepaths and display in a graph
- display the top N slowest =unique= purepaths
I'm about ready to pull out my last grey hair.
Help!
Thanks!!
Thanks, yes it appears the business transaction is the best way to go at the moment.
It will slice up the transactions by URL and give you counts and some nice aggs which you can then sort and drill into...
Thanks!
I almost thought that thats what you wanted. Unfortunately this is currently not supported. However - we are investing heavily in our new Dynatrace Web UI where we wiill support exactly these use cases to answer these "Top X" questions. Stay tuned on whats coming.
For the current versions out there you have use the approaches I highlighted. You could create that BT you mentioned and then just look at the Top 10 for instance. Not as convenient - but - would work
Thanks Andi.
I see the Top X filter but I want to show the top X unique purepaths - using that filter simply shows me ten of the SAME poorly performing purepaths. 
I know this is a little different way to look at the data, but it's useful, none the less.
Hi Don
What makes a PurePath unique for you? The URL? If thats what it is you can create a Business Transaction that splits by URL Pattern. But - this approach has its risk as well - especially if you have A LOT OF unique URLs. Because that is then creating a lot of measures in the performance warehouse for each indivdiual URL which you dont need. It will fill up your PWH and will actually stop your BT from working after you reached a certain "max split number". But - if the number is reasonable, e.g: < 10k - then this would be an option
For your #2: You could open the PurePath dashlet and specify the "Top X" Filter for PurePath Execution time. This will then show you the slowest PurePaths in the current timeframe. Here be aware that you should pick a timeframe that is reasonable. Something like last 6 hours is ok. Something like last 30 days might be an overkill
Andi
