Log Analytics is the central warehouse for all logs from your VNets, VMs, SQL databases, and App Services. To query them, you need to master KQL (Kusto Query Language).
KQL is designed to be readable and easy to pipe. It feels like a mix of SQL and PowerShell.
requests | where success == false | summarize count() by bin(timestamp, 1h) | render timechart.
In one line, you've found all failed requests, grouped them by hour, and visualized them in a chart.
You can save your KQL queries into **Azure Workbooks**. These are interactive, visual dashboards that your whole team can use to monitor the health of your .NET ecosystem. You can even combine data from multiple subscriptions into a single view.
Q: "Is KQL hard to learn?"
Architect Answer: "If you know SQL, you'll learn the basics of KQL in 10 minutes. The biggest shift is the 'Pipe' (|) operator. The data flows from top to bottom, being filtered and reshaped at every step. It is significantly faster than SQL for searching millions of rows of text data."