The integrated monitoring tools provided tremendous value when observing, analyzing, and making performance recommendations because of the various levels of detail they provided. We could observe all the key SQL tuning metrics – including waits, cache, transactions, memory, and many more. For more information about monitoring SQL with Grafana, see the Grafana documentation.
The following figures were captured from Grafana during the 44-minute run of the Text-6 configuration. The SQL Managed Instance metrics screenshots were taken for SQLMI-10. We could clearly observe increasing load as we added virtual users. We found it simple to interpret the results because all DBA-centric metrics can be found on a single screen.
Figure 27. Test-6 SQL metrics
Figure 28. Test-6 SQL metrics continued
Figure 29. Test-6 container metrics
Figure 30. Test-6 pod metrics
Figure 31. Test-6 worker node CPU metrics
We also leveraged Azure Monitor Container Insights. This service can monitor the entire AKS-HCI deployment. The following figures were also captured during the 44-minute run for the Test-6 configuration. This was an effective tool for cluster, node, controller, and container level metrics altogether in one simple interface. This service that is combined with Grafana at the data controller level provided all the pertinent information that we needed to make timely performance-related decisions.
Figure 32. Test-6 CPU usage by container
Figure 33. Test-6 node usage