Can you let me know what measure the default incident: Response time degraded for slow requests is using? I
tried creating my own and was able to recreate an alert very similar,
but it was missing the Business Transaction detail that's included in the default alert.
You can write your own description in the business transaction and also when you create the incident so when you will receive the incident via email that description will be there.
The business transaction looks like more than just a description. I am attaching a picture. When we click on the link in the default incident to open in dynatrace we are taken to the incident. This is helpful to diagnosis quicker and the spot we are taken to correlates with the business incident listed in the attachment. In my custom incident, there is not business transaction and when you click on open in dynatrace you are take to the applicationdynatrace-applications-screenshot-3.jpg and you have to drill down to find the process that triggered the alert.
The reason the Business Transaction detail is missing is because that alert is from the out of the box business transaction, "Web Page Requests" which is a baselined Business Transaction. You would have to create a custom Business Transaction and add it to the list of baselined BT's from the "Configure" option on the application overview screen, which can be navigated to by clicking the "Applications" tab from the main monitoring overview (The main screen when you first open the client). Let me know if you have any questions on how to set that up. Thanks!
I think I was able to complete the steps you mentioned. The alert does include some business transaction data, however when I click on open in dynatrace I am not taken to the issue. Have a look at the two screenshots I am attaching. One screenshot shows the e-mail alert I received where it notes a request time of 1820.47 ms. The second screenshot shows the screen I receive when I click on open in dynatrace. Here you will see I can't locate the time when the response time was 1820.47 ms. I tried scrolling 10s forward and back and don't see it. In the default alert this appears to be working. dynatrace-applications-screenshot-4.jpgdynatrace-applications-screenshot-5.jpg
Can you change the splitting options of your custom created business transaction from 5% to No Restriction for sometimes?
Check the violated PurePaths on the next incident.
Would you mind to share the configuration of incident?
A couple things to point out here. You can see in the email you received that response time degraded for 'slowest 10% of requests" which is taking into account baseline data. When you view the business transaction in the 10 second timeframe, you are only viewing the transactions that took place during that time. but the alert is telling you that for a 10 second period, the slowest 10% (baseline) of the business transaction requests were at 1820.47 ms for significant measures during that time.
You can also see that the splitting in the email is <...> which means this is an aggregation of all other web requests that were not made into their own splitting, as mentioned above, you can change the splitting options to your liking in the business transaction editor.
If you increase the timeframe you are looking at, you'll see that the splitting <...> was over 1000 ms at some point.
You can also easily view response time degradation for baselines easily from the application overview I mentioned earlier, you can find more details about it here:
Please let me know if you have further questions. Thanks!
Let me ask one more question and than I think I have enough information to achieve the base alerts we want to get started with.
I would like to create custom incidents by application or agent group for the following default incidents, can you let me know what measures to use. I can't seem to locate them.
Application Process Unavailable - Agent connection lost
Application Process Out-of-Memory - Out-of-Memory on Agent
Application Process Unhealthy - GC health of process