• TRAFFIC
  • RESOURCES
  • TIMING
  • HUMAN

How to analyze performance spikes with your AI

Connect your Azure Log Analytics Workspace once. From there, your AI handles the log scanning and counting for you.

Ask AI about this page

Short answer

How can my AI find performance spikes in Azure?

Your AI uses the connector to run queries that count log entries over specific time windows. This helps you spot sudden traffic surges or resource exhaustion. You'll get a clear view of when and where the spikes happened.

Where the data lands.

The results of your analysis.

The AI processes your telemetry and organizes the findings into these outcomes.

TRAFFIC

Detect surges

The AI counts request logs to show you exactly when traffic volume jumped. You'll see the specific time windows where the load increased.

RESOURCES

Find exhaustion

The AI identifies logs tied to CPU or memory limits. This shows you which specific resources are hitting their ceiling.

TIMING

Isolate windows

The AI pinpoints the exact timestamps of log frequency changes. You get the precise moments the system behavior shifted.

HUMAN

Manual triage

The AI flags the most relevant log segments for your review. You decide if these require a code fix or a configuration change.

The workflow

What your AI does when the logs arrive.

The AI acts as the interface between your questions and your Azure telemetry.

  1. Fetch telemetry

    Your AI pulls the necessary data from your Azure Log Analytics Workspace.

    query_logs

  2. Aggregate counts

    The AI counts log entries over the time windows you define to find patterns.

    query_logs

  3. Identify anomalies

    The AI compares log volumes to find where the counts deviate from the norm.

    query_logs

  4. Report findings

    The AI summarizes the log counts into a plain text explanation of the spike.

    query_logs

Try it

Copy these to start.

Use these prompts to begin your investigation.

Starting points

Replace the workspace names and timeframes with your own specific details.

Azure Log Analytics Workspace Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_h8mzxcilFYlijQMfr5u3GIGk82xs7W4r1G5Um6EK/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Azure Log Analytics Workspace capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "azure-log-analytics-workspace-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_h8mzxcilFYlijQMfr5u3GIGk82xs7W4r1G5Um6EK/mcp"
    }
  }
}
Copy into chat04

  • Count the error logs in my 'Prod-Web-App' workspace for the last 30 minutes.

  • Show me the count of request logs in 'Azure-Gateway' every 5 minutes for the last 2 hours.

  • Did my 'DB-Cluster' workspace see a spike in CPU exhaustion logs between 2 PM and 4 PM today?

  • Compare the log volume for 'API-Service' in the last hour versus the hour before that.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Azure Log Analytics Workspace once, then ask.

Follow the link to authorize the connection. Your credentials stay encrypted on our side, and you can start querying immediately.

Connect Azure Log Analytics Workspace to your AI

FAQ

How this task behaves.

  • 01

    Can the AI change my Azure settings?

    No. The AI can only read and query logs. It cannot modify your Azure resources or change any configurations.

  • 02

    Does the AI delete any of my log data?

    No. The AI only has read access through the query tool. It cannot delete or alter your logs.

  • 03

    Can it alert me in real-time?

    No. The AI performs analysis based on the queries you ask. It does not run in the background as a continuous monitoring service.

  • 04

    How does it handle large volumes of logs?

    The AI uses the query tool to aggregate data into counts, so it doesn't need to read every single line of raw text to give you an answer.

  • 05

    Can it fix the resource exhaustion it finds?

    No. The AI identifies the problem, but you have to perform the actual fix in your Azure environment.

  • More questions about Azure Log Analytics Workspace? The Connector page answers them. See everything the Azure Log Analytics Workspace Connector can do

Connect Azure Log Analytics Workspace to Claude, Cursor, ChatGPT & more