Skip to content

Checklists

Fast triage guides for the first 10 minutes of an Azure Functions investigation.

These checklists help you quickly narrow down the problem category and identify which playbook to follow for deeper analysis.

graph TD
    A[Initial incident signal] --> B{Choose first-response checklist}
    B --> C[Triggers Not Firing checklist]
    B --> D[High Latency checklist]
    B --> E[Scaling Issues checklist]
    C --> F[Trigger playbooks]
    D --> G[Performance playbooks]
    E --> H[Scaling playbooks]
Checklist When to Use
Triggers Not Firing Functions not executing, zero invocations, trigger listener failures
High Latency Slow responses, elevated P95, timeout errors
Scaling Issues Queue backlog growing, executions flat, cold start spikes

Portal Walkthrough

Before diving into CLI and KQL, use these Portal blades to get a quick visual overview of the Function App state.

Function App Overview

[Observed] The Overview blade shows the app status (Running), hosting plan (KoreaCentralLinuxDynamicPlan (Y1: 0)), runtime version (4.1048.200.26180), and the list of deployed functions with their trigger types and enabled status:

Function App overview blade showing 2 HTTP-triggered functions (health, hello) with Enabled status

[Inferred] If the Status field shows anything other than Running, or if functions show Disabled, that is the first signal to investigate. The "Migrate your app to Flex Consumption" banner indicates this is a Linux Consumption plan approaching EOL.

Log Stream

[Observed] The Log stream blade connects to the live output of the Function App host. The status Connected! confirms the host is running and reachable:

Log stream blade showing Connected status

[Inferred] If Log stream fails to connect or shows no output for more than 30 seconds, the app may be scaled to zero (Consumption) or the host may be in a crash loop. Proceed to the Triggers Not Firing checklist.

Diagnose and Solve Problems

[Observed] The Diagnose and solve problems blade shows active risk alerts and categorized troubleshooting tools:

Diagnose and solve problems with Availability warning and 5 troubleshooting categories

[Inferred] The Availability: 1 Warning alert is a quick indicator that the platform has detected an issue. Click View more details to see the specific detector findings before running manual KQL queries.

Common first actions

Regardless of which checklist you follow, always start with these three checks:

  1. Azure Service Health — Rule out regional platform issues
  2. Application Insights Live Metrics — Confirm whether failures are active
  3. Recent deployments — Check if anything changed in the incident window

See Also

Sources