AutoGineer
Generated banner for Azure Diagnostics
Generated logo for Azure Diagnostics

Azure Diagnostics

MicrosoftOfficialCreated by Microsoft

  • Skill
  • Workflow
  • Cloud and DevOps
  • MIT
  • Low risk
  • Has scripts
Download ZIP

From the author

Debug Azure production issues using AppLens, Azure Monitor, resource health, and safe triage, including evidence-only diagnosis without live access. WHEN: debug production issues, troubleshoot app service, app service high CPU, app service deployment failure, troubleshoot container apps, container app 503 or target port mismatch, troubleshoot functions, function telemetry missing or stopped after a config/connection-string change, troubleshoot AKS, VM RDP, Linux SSH, VM black screen, can't connect to VM, reset VM password, NSG or firewall blocking, kubectl cannot connect, kube-system/CoreDNS failures, pod pending, crashloop, node not ready, upgrade failures, analyze logs, KQL, insights, image pull failures, cold start issues, health probe failures, resource health, root cause of errors, troubleshoot event hubs, troubleshoot service bus, messaging SDK error, AMQP connection failure, message lock lost, lock renewal during long-running processing, service bus dead letter.
— the description by Microsoft

AutoGineer summary

Written by AutoGineer from the skill’s own files, not by the author

Azure Diagnostics provides a structured, authoritative approach to investigating and resolving production issues across various Azure services. It emphasizes a systematic flow: identifying symptoms, checking resource health, reviewing logs, analyzing metrics, and investigating recent changes. The skill supports both live troubleshooting via Azure CLI and AppLens, as well as evidence-only analysis when live access is restricted. It includes specialized guidance for App Service, Container Apps, Azure Functions, AKS, VMs, and Messaging services, ensuring consistent root-cause analysis.

What it does

  • Provides a standardized diagnostic workflow for Azure production environments.
  • Integrates with AppLens and Azure Monitor for automated issue detection and log analysis.
  • Offers service-specific troubleshooting paths for App Service, AKS, VMs, and more.
  • Supports evidence-only diagnosis for secure, offline troubleshooting scenarios.
  • Includes KQL query patterns for deep-dive log and resource health analysis.
  • Facilitates the use of bundled diagnostic scripts for efficient data collection.

When to use it

  • When investigating production incidents like high CPU, deployment failures, or service crashes.
  • When troubleshooting connectivity issues such as RDP/SSH failures or network blocking.
  • When diagnosing Kubernetes-specific issues like pod crashloops or image pull failures.
  • When analyzing missing telemetry or messaging errors in Event Hubs and Service Bus.
  • When performing root-cause analysis on unhealthy or degraded Azure resources.

Advantages

  • Reduces mean time to resolution (MTTR) with a clear, step-by-step diagnostic process.
  • Ensures consistent troubleshooting practices across different Azure service types.
  • Supports secure workflows by allowing diagnosis without requiring live environment access.
  • Leverages powerful built-in tools like AppLens and Azure Resource Graph for rapid insights.
  • Provides actionable remediation steps based on identified symptoms.

Usage examples

“Troubleshoot my App Service that is experiencing high CPU usage.”
The agent will initiate the App Service diagnostic flow, run the appservice-diagnostics script, check metrics, and guide you through profiling via Kudu.
“Why is my Azure VM unreachable via SSH?”
The agent will follow the VM connectivity troubleshooting guide, checking NSG rules, firewall status, and VM agent health.
“Analyze these logs to find why my Container App is failing to pull the image.”
The agent will perform an evidence-only analysis of the provided logs to identify image pull errors or registry authentication issues.
“Check the health of my AKS cluster and identify why pods are in a crashloop.”
The agent will check resource health, query logs for the affected pods, and route to the AKS-specific troubleshooting documentation.

Install

  1. Download the skill as a ZIP:

    https://mp-staging.autogineer.com/agent-skills/microsoft/azure-diagnostics/download

  2. Unzip it into the cross-client skills folder of your project:

    mkdir -p .agents/skills && unzip azure-diagnostics.zip -d .agents/skills

Documentation by Microsoft

Azure Diagnostics

AUTHORITATIVE GUIDANCE — MANDATORY COMPLIANCE

This document is the official source for debugging and troubleshooting Azure production issues. Follow these instructions to diagnose and resolve common Azure service problems systematically.

Triggers

Activate this skill when user wants to:

  • Debug or troubleshoot production issues

  • Diagnose errors in Azure services

  • Analyze application logs or metrics

  • Fix image pull, cold start, or health probe issues

  • Investigate why Azure resources are failing

  • Find root cause of application errors

  • Troubleshoot App Service issues (high CPU, deployment failures, crashes, slow responses, TLS/custom domains)

  • Respond to prompts like "troubleshoot app service", "app service high CPU", or "app service deployment failure"

  • Troubleshoot Azure Function Apps (invocation failures, timeouts, binding errors)

  • Find the App Insights or Log Analytics workspace linked to a Function App

  • Troubleshoot AKS clusters, nodes, pods, ingress, or Kubernetes networking issues

  • Troubleshoot Azure VM connectivity issues (RDP/SSH failures, port 3389/22 timeouts, NSG or firewall blocking, credential resets)

  • Troubleshoot Azure Messaging SDK issues (Event Hubs, Service Bus connection failures, AMQP errors, message lock issues)

  • Root-cause a symptom that looks like a healthy-runtime-but-wrong-destination problem (e.g., telemetry/messages stop flowing to the expected resource after a config change) — invoke this skill for the matrix even when MCP tools could answer the surface question alone

Rules

  1. Start with systematic diagnosis flow

  2. Use supplied evidence first. If the user requests evidence-only analysis or prohibits live Azure access, do not query or modify Azure. Diagnose from the supplied logs, metrics, and timeline; identify assumptions and provide optional verification steps.

  3. Use AppLens (MCP) for AI-powered diagnostics when available

  4. Check resource health before deep-diving into logs

  5. Select appropriate troubleshooting guide based on service type

  6. Document findings and attempted remediation steps

  7. Prefer bundled diagnostic scripts over ad hoc calls for permitted live collection when identifiers are known. If supplied evidence suffices or live access is prohibited, don't run them; otherwise run the applicable script first.

  8. Route AKS incidents to the dedicated AKS troubleshooting document

  9. Keep baseline AKS diagnostics available even when the optional aks-skills add-on is absent, declined, or unsupported by the host

Quick Diagnosis Flow

  1. Identify symptoms - What's failing?

  2. Check resource health - Is Azure healthy?

  3. Review logs - What do logs show?

  4. Analyze metrics - Performance patterns?

  5. Investigate recent changes - What changed?

Troubleshooting Guides by Service

ServiceCommon IssuesReference
Container AppsImage pull failures, cold starts, health probes, port mismatchescontainer-apps/
App ServiceHigh CPU, deployment failures, crashes, slow responses, TLS/custom domainsapp-service/
Function AppsApp details, invocation failures, timeouts, binding errors, cold starts, missing app settingsfunctions/
AKSCluster access, nodes, kube-system, scheduling, crash loops, ingress, DNS, upgradesAKS Troubleshooting
ComputeVM RDP/SSH connectivity, NSG/firewall blocks, credential resets, VM agent/tooling issuesVM Connectivity Troubleshooting
MessagingEvent Hubs & Service Bus SDK errors, AMQP failures, message lock, connectivityMessaging Troubleshooting

Routing

  • Keep Container Apps and Function Apps diagnostics in this parent skill.

  • Route active AKS incidents, AKS-specific intake, evidence gathering, and remediation guidance to AKS Troubleshooting.

  • Route Azure VM RDP/SSH connectivity, NSG/firewall, credential reset, and VM agent troubleshooting to VM Connectivity Troubleshooting.

  • Route Azure Messaging SDK troubleshooting (Event Hubs, Service Bus) to Messaging Troubleshooting.

Quick Reference

Common Diagnostic Commands
# Check resource health
az resource show --ids RESOURCE_ID
# View activity log
az monitor activity-log list -g RG --max-events 20
# Container Apps logs
az containerapp logs show --name APP -g RG --follow
# Function App logs (query App Insights traces)
az monitor app-insights query --apps APP-INSIGHTS -g RG \
  --analytics-query "traces | where timestamp > ago(1h) | order by timestamp desc | take 50"
AppLens (MCP Tools)

For AI-powered diagnostics, use:

mcp_azure_mcp_applens
  intent: "diagnose issues with <resource-name>"
  command: "diagnose"
  parameters:
    resourceId: "<resource-id>"

Provides:
- Automated issue detection
- Root cause analysis
- Remediation recommendations
Azure Monitor (MCP Tools)

For querying logs and metrics:

mcp_azure_mcp_monitor
  intent: "query logs for <resource-name>"
  command: "logs_query"
  parameters:
    workspaceId: "<workspace-id>"
    query: "<KQL-query>"

See kql-queries.md for common diagnostic queries.

Check Azure Resource Health

Using MCP
mcp_azure_mcp_resourcehealth
  intent: "check health status of <resource-name>"
  command: "get"
  parameters:
    resourceId: "<resource-id>"
Using CLI
# Check specific resource health
az resource show --ids RESOURCE_ID

# Check recent activity
az monitor activity-log list -g RG --max-events 20

References

Files

troubleshooting/aks/pod-failures.md

Pod Failures & Application Issues

Evidence Bundle Script

For any pod symptom below, run the pod-evidence script to collect the same read-only evidence bundle. Per pod it digests STATUS, STATE (exit code, reason, last state), EVENTS, current/previous LOGS, and RESOURCES (requests vs top). It only gathers; interpret with the tables.

Bash ../../scripts/pod-evidence.sh · PowerShell ../../scripts/pod-evidence.ps1

../../scripts/pod-evidence.sh <pod-name> -n <namespace>   # one pod
../../scripts/pod-evidence.sh --all-failing               # all unhealthy pods

PowerShell: ../../scripts/pod-evidence.ps1 <pod-name> -Namespace <namespace> (-AllFailing scans all).

CrashLoopBackOff

Pod starts, crashes, restarts with exponential backoff (10s, 20s, 40s... up to 5m).

Diagnostics: pod-evidence → read STATE (exit code, reason, last state) and PREV LOGS (last crashed container).

Decision tree:

Exit CodeMeaningFix Path
0App exited successfully (unexpected for long-running)Check if entrypoint/command is correct; app may be a one-shot
1Application errorRead logs - unhandled exception, missing config, bad startup
137OOMKilled (SIGKILL)Increase resources.limits.memory; check for memory leaks
139Segfault (SIGSEGV)Binary compatibility issue or native code bug
143SIGTERM - graceful shutdownPod was terminated; check if liveness probe killed it

OOMKilled specifically: the STATE section shows terminated=OOMKilled and RESOURCES shows the memory limit vs live usage. Fix: increase resources.limits.memory or optimize application memory usage.

OOM kill tracing with Inspektor Gadget: Run trace_oomkill for the pod to see which process was killed and memory at kill time: scripts/run-ig.sh --gadget trace_oomkill --pod <pod-name> --ns <namespace> (or run-ig.ps1).

Deep diagnostics with Inspektor Gadget (when logs and describe are inconclusive):

Use scripts/run-ig.sh (or run-ig.ps1) with --pod <pod-name> --ns <namespace> and these gadgets:

  • trace_exec — see what the container executes at startup

  • trace_open — find missing configs/secrets (retval -2 = ENOENT, -13 = EACCES)

  • snapshot_process — list running processes in the pod

ImagePullBackOff

Pod can't pull the container image.

Diagnostics: pod-evidence → read EVENTS for the exact pull error.

Error MessageCauseFix
ErrImagePull / ImagePullBackOffImage name or tag is wrongVerify image name and tag exist in the registry
unauthorized: authentication requiredMissing or wrong pull secretCreate/update imagePullSecrets on the pod or service account
manifest unknownTag doesn't existCheck available tags in the registry
context deadline exceededRegistry unreachableCheck network/firewall; for ACR, verify AKS -> ACR integration

ACR integration check:

# Verify AKS is attached to ACR
az aks check-acr -g <rg> -n <cluster> --acr <acr-name>.azurecr.io

Pending Pods

Pod stays in Pending - scheduler can't place it.

Diagnostics: pod-evidence → read EVENTS for why scheduling failed.

Event MessageCauseFix
Insufficient cpu / Insufficient memoryNo node has enough resourcesScale node pool; reduce resource requests; check for overcommit
node(s) had taint ... that the pod didn't tolerateTaint/toleration mismatchAdd matching toleration or use a different node pool
node(s) didn't match Pod's node affinity/selectorAffinity rule unsatisfiableCheck nodeSelector or nodeAffinity rules
persistentvolumeclaim ... not found / unboundPVC not readyCheck PVC status; verify storage class exists
0/N nodes are available: N node(s) had volume node affinity conflictZonal disk vs pod in different zoneUse ZRS storage class or ensure same zone

Readiness & Liveness Probe Failures

Readiness probe failure -> pod removed from Service endpoints (no traffic). Liveness probe failure -> pod killed and restarted.

Diagnostics: pod-evidence → EVENTS shows Readiness/Liveness probe failed; STATUS shows the READY column (must be n/n).

SymptomCauseFix
READY shows 0/1 but pod is RunningReadiness probe failingCheck probe path, port, and app health endpoint
Pod restarts repeatedlyLiveness probe failingIncrease initialDelaySeconds; check if app starts slowly
Probe timeout errorsApp responds too slowlyIncrease timeoutSeconds; check app performance

💡 Tip: Set initialDelaySeconds on liveness probes to be longer than your app's startup time. A common mistake is killing pods before they finish initializing.

Resource Constraints (CPU/Memory)

Check actual usage vs limits: pod-evidence → RESOURCES compares requests/limits against live top usage. To rank a namespace by memory: kubectl top pod -n <namespace> --sort-by=memory.

SymptomCauseFix
OOMKilled (exit code 137)Container exceeded memory limitIncrease limits.memory or fix memory leak
CPU throttling (slow responses)Container hitting CPU limitIncrease limits.cpu or remove CPU limits
Pending - insufficient resourcesRequests exceed available node capacityLower requests, scale nodes, or use larger VM sizes

⚠️ Warning: Setting CPU limits can cause unnecessary throttling even when the node has spare capacity. Many teams set CPU requests but not limits. Memory limits should always be set.

Requirements and compatibility

Works with: Claude Code, Claude apps, OpenAI Codex, GitHub Copilot, Cursor, Gemini CLI, Kiro, OpenCode, Goose, Amp, Roo Code, JetBrains Junie, Factory Droid, Devin (ex-Windsurf), Mistral Vibe, OpenHands, Letta

“Works with” means the tool reads the open Agent Skills format; a skill written for one tool may use features others lack.

Needs: Needs a command-line tool

Security

What our scanner noticed· Low risk

  • Instructions that use the network
  • Scripts that your agent may run

This skill includes scripts that your agent may run. Read them before you use it.

A static scan is not a guarantee. Skills are third-party content: review a skill before you let an agent use it.

Attribution and licence

Created by
Microsoft
Publisher
Microsoft (Official)
Source
microsoft/skills · .github/plugins/azure-skills/skills/azure-diagnostics · commit 0696600
Licence
MIT License (redistribution allowed)
  • LICENSE (added by AutoGineer from .github/plugins/azure-skills/LICENSE)
  • ATTRIBUTION.md (written by AutoGineer)
Synced
Also distributed in
microsoft/azure-skills

AutoGineer is not affiliated with or endorsed by Microsoft.

Download ZIP

The ZIP holds the skill’s files, its licence files and an ATTRIBUTION.md. Download azure-diagnostics.zip