I'm investigating the enhanced possibilities of data viz using Tableau. I've set up a test sql server perf warehouse for appmon and have Tableau connecting successfully. But the question is actually which tables are holding the data.
For example the measure_[low|mid|high] and dynamic_measure and particularly measure_temp[1-2] look promising (when used with appropriate joins to the measure_id... etc).
Has anyone used the perf warehouse db directly via sql, as this is a similar process.
Solved! Go to Solution.
our schema isn't designed with direct access in mind, but it's relatively simple.
The measurement_[high|mid|low] tables contain measure points in 1 min / 1 hour / 1 day resolution.
The timestamp is unix epoch time in UTC in milliseconds.
The measureid points to the dynamic_measure.id column.
The dynamic_measure table contains information about the specific "splitting" that a measurement was booked for. It contains pointers into the source (src_id), application (application_id), systemprofile (system_id), metric (metric_id) and metricgroup (metric_group_id) look-up tables.
The other columns are for internal use.
You can also use the threshold_high table to read out thresholds for the measures.
This set of tables contains the measured data that we collect from purepaths and host / process health etc metrics.
There are other types of data contained in the pwh:
- incident records
- test integration results (like unit test runs or some load driver runs etc)
- baseline data
They're probably a bit more tricky to read out and interpret.
In general, I recommend creating a read-only user for work on our schema. Let me know if you need more details, cheers,