Experts, I am trying to implement in new Dynatrace similar to how we did in Appmon for root-cause analysis. (sorry for such a long question, pls see below)
Our requirement is, build a Dashboard for all Business Transactions and should be able to drill down to all tiers from the chart for an Account ID. I appreciate your suggestions on how to achieve this.
New DT Dashboard Requirement(Steps 1-6) and how I achieved this in Appmon(steps7-8)->
- Capture all Key Transactions 95th or Average Response times and Throughput (requests count) in Dashboard.
- Do the Splitting on the Transactions by Account ID [Account ID can be found in the requested URL, in Appmon captured this by creating a measure]
- From the Dashboard/Chart, if I want to find the requests related to an Account ID, I can simply search that Account ID from the chart and apply the filter on it so that I would be viewing only requests related to that particular Account ID.
- After #3, want to click on a request to see the contribution of the client and Serverside.
- from #4, if I notice the slowness is due to Server-side, I should be able to see the Purepaths related to that particular request to find the root cause.
- from #4, if the slowness is due to the user action side, then I should be able to view the Browser clicks or User Actions that are captured under User Sessions or Applications to find the root-cause for slowness on client-side.
- In Appmon, how I achieved this was by creating Two types of measures [one for Account ID and other for the Request pattern] and created Two types of Business Transactions.
- Business Transaction for Entire Round-trip response time – User Action Duration [Estimated Client time + Network Contribution + Server Contribution]
- Business Transaction for Server-Side Contribution – Purepath response time only
- Once the Splitting is created, I have an option in Appmon while adding BT to the dashboard to do the Split or No split to plot Aggregated values of all accounts on that BT.
- Splitting by Account ID -> with this I can do a search on Account ID in the chart to do down on the requests related to that Account ID.
- While Adding this BT with No Splitting – Gives Aggregate value of all Account ids’ for that business transaction.
For Above Screenshot#
- Business Transaction Name
- Measure to do the Filter (abc-logins)
- Calculate Results for User Actions– Measures used available in Appmon [Estimated Client Time, Network Contribution, and Server Contribution]; For Server-side BT, do not select any so that it gives only Purepath response times.
- Split Results – Created a measure for Account ID [AccountID], and Application.
I understand that, in the new Dynatrace, with AI it auto-detects the slow performing transaction or the transaction which breaches the SLA and shows it in Problems Dashlet and eventually sends the alerts. For the sake of designing dashboards, I want to bypass this option for now.
Thanks in advance for your help!
Hi @Shuresh k.
In dynatrace things are managed a bit in different way than in AppMon as you already noticed.
Will try to point the direction to answers for your steps.
1. Multi Dimensional Analysis (aka: MDA) and create metrics if needed
2. Request Attribute (aka: RA) on the URL and URL query
3. You will find the answer under user action contributors breakdown
4. In order to check the web requests of the user action you will need to go to the waterfall chart
there you can see the median time of each web request
5. Switch to the instances of this action tab and drill down to a PP
6. See answer for step 4.
7. See answer for step 3.
8. RA will give you the ability to split web request for example
HTH to clear a bit how to look on dynatrace data
Yos
thanks @Chad T. I am trying to do a custom chart and eventually pin in a dashboard. So, what is the metric need to be selected here?
Featured Posts
