Can we compare response time, error rate (count), availability of services and other similar metrics for this hour to last 1 hour data ? If Yes, could you please help me out... I am looking to build a dashboard with the comparison.
I want to know how it is done for DCRUM and dynatrace both.
Solved! Go to Solution.
Hello @Raviteja E.
Yes. You should be able to compare the response time, error and other desired metrics available in the Dynatrace/AppMon and for this you will have to create a custom measure for each transaction.
Use the Functional Health dashlet to analyze the transaction load and failure rate compared to another time frame for a whole application.
In the DCRUM/NAM OOB dashboards are available to compare the Availability, Performance etc... with the baseline (last 9 days) for each software service.
Yes, you can compare virtually any metric between different time ranges within the same section.
Dynatrace NAM (DC RUM) is very versatile when it comes to reporting capabilities and custom analytics you can do on the data gathered by NAM.
Since NAM 2018, not only you can precisely define those time ranges to be actually anything you can choose from calendar (previously you could only define relative time ranges, i.e., shifted by 1 hour or 2 days, etc.) but you can also use calculated columns to take your analysis to another level.
Here is an example of a report when I pulled Operation time from a single monitoring interval, an hour before that, they whole day until that moment, the whole day yesterday, (aggregated over) last 5 days, and operation time calculated from baseline (by default calculated for last 10 days):
Once you have those measurements (in separate columns, delivered from separate benchmark data definitions) you can build in analytics into your report:
You can see this report in action at https://apm.gdn.demo.dynatrace.com/LSServlet?lsEn... (guest / guest)
I have attached definition of this report to this post.comparing-metrics-from-different-timesxml.zip
Btw. I left the "filter" information on actual time ranges rendered for each benchmark data set. Usually this is not necessary and this information can be hidden from end user, and requested to be rendered only in PDF.
I would say almost. I have extended the original report with a second section that does not use calculated columns (which are a new thing in NAM 2018). I have added one more comparison coloring based on Baseline absolute, which colors based on thresholds expressed in actual differences in values rather than their relative differences. I also added tooltips to both tables.
If you want to have columns with calculated differences or icons, you will need NAM 2018. And I would strongly encourage you to upgrade to NAM 2018 (SP2 coming by the end of the year), which comes with many more new goodies 🙂
I am unable to create hours ago column, Today and yesterday column in the report i am generating...
This is how i am creating my report :
Select DMI--> new report --> under dimension and metrics --> selecting software service and operation time respectively.
Next --> Add benchmark and in data view --> selected software service,operations and site baselines to create baseline data. Add this to my report i could pull down only 2 columns one with Operation time (current time) and other operation time (baseline time)...
How can i add hours ago column, Today column and also i am not sure of custom time range how to add it to the same report.. If possible can you help me in steps how to create the above report. I am little new to this...
Yes, you are right. I'm sorry, I forgot we also added one more thing to NAM 2018 - time ranges set per benchmark, not only per section.
In DC RUM 2017 May you can still have relative time range, but it will only be calculated for the very same length, e.g., if you show section for last 1 hour, you can show benchmark data for 1 hour but starting couple of hours or couple of days ago:
The trick is to add Time dimension to Benchmark data source and define the shift.This way you can still compare different time ranges but they need to have the same length. I'm afraid I don't have any DC RUM 2017 setup available for the outside world to show how the report would look like. The screenshot above comes from a 2017 setup.