Screenshot from the article

Introduction

Amazon DevOps Guru is an AI-powered operations service that helps identify and fix issues in applications running on AWS. It continuously monitors infrastructure metrics, logs, and events to detect anomalies and recommend actionable insights to improve performance and reliability.

This blog provides an in-depth guide on how to set up, configure, and use Amazon DevOps Guru effectively with real-world examples, advanced scenarios, and FAQs to help DevOps engineers enhance their observability and incident resolution workflows.

🚀 Why Use Amazon DevOps Guru?

🔹 Key Benefits

  • Proactive Issue Detection: Uses ML to detect performance bottlenecks and anomalies.
  • Automated Insights: Provides actionable recommendations to mitigate risks.
  • Seamless AWS Integration: Works across AWS services like EC2, RDS, Lambda, and Kubernetes.
  • Cost Optimization: Helps identify underutilized resources and inefficiencies.

📌 Use Cases

  • Preventing Outages: Detect unusual CPU spikes in EC2 before an outage occurs.
  • Troubleshooting Slow Database Queries: Identify slow RDS queries affecting application response time.
  • Optimizing Lambda Performance: Detect memory allocation issues in serverless applications.
  • Improving Kubernetes Reliability: Spot failing pods in Amazon EKS clusters before they impact users.

🛠 Setting Up Amazon DevOps Guru

✅ Prerequisites

  1. AWS Account with admin privileges.
  2. IAM Permissions to manage DevOps Guru (AmazonDevOpsGuruFullAccess policy).
  3. Active AWS Services (EC2, RDS, Lambda, ECS, EKS, etc.) to monitor.

🔧 Step-by-Step Setup

Step 1: Enable DevOps Guru

aws devops-guru update-account-settings --resource-coverage {COVERAGE}

Alternatively, enable it via AWS Console:

  • Navigate to AWS DevOps Guru → Click Enable DevOps Guru.
  • Choose Resource Coverage Scope (specific services or all AWS resources).

Step 2: Define Analysis Scope

You can configure DevOps Guru to monitor:

  • Entire AWS Account
  • Specific AWS Services (EC2, RDS, Lambda, etc.)
  • Tags & Resource Groups

Example CLI command to set up EC2 monitoring:

aws devops-guru update-resource-collection --resource-collection {"CloudFormation": [{"StackNames": ["MyAppStack"]}]}

Step 3: Configure Notification Channels

To send alerts via SNS (Slack, Email, PagerDuty):

aws sns create-topic --name DevOpsGuruAlerts
aws sns subscribe --topic-arn arn:aws:sns:us-east-1:123456789012:DevOpsGuruAlerts --protocol email --notification-endpoint you@example.com

Step 4: Review Findings & Insights

Once DevOps Guru is enabled, it starts analyzing logs, events, and metrics. Navigate to:

  • AWS Console → DevOps Guru Dashboard to see insights.
  • Run CLI command to fetch insights:
aws devops-guru list-insights --status-filter {"Ongoing":{}}

🏆 Advanced Real-World Scenarios

1️⃣ Detecting High Latency in an RDS Database

Scenario: A production database in Amazon RDS (PostgreSQL) experiences intermittent slow queries.

📌 How DevOps Guru Helps:

  • Detects Anomalies: Notifies high query execution times.
  • Root Cause Analysis: Shows related EC2 instances under heavy load.
  • Recommended Action: Suggests increasing max_connections or optimizing queries.

👉 Example AWS CLI to View Insights:

aws devops-guru describe-insight --id insight-12345

2️⃣ Identifying Memory Leaks in Lambda Functions

Scenario: A Lambda function used in a serverless app suddenly runs out of memory and fails.

📌 How DevOps Guru Helps:

  • Detects Memory Spikes: Flags sudden memory exhaustion.
  • Root Cause Analysis: Shows a loop consuming excess memory.
  • Recommended Action: Suggests increasing memory allocation and optimizing recursion depth.

👉 Check Lambda Anomalies:

aws devops-guru list-anomalies-for-insight --insight-id insight-67890

3️⃣ Troubleshooting Failing Kubernetes Pods in Amazon EKS

Scenario: A Kubernetes cluster in Amazon EKS sees pod restarts due to high CPU.

📌 How DevOps Guru Helps:

  • Detects Pod Failures: Identifies pods restarting frequently.
  • Root Cause Analysis: Detects CPU throttling in worker nodes.
  • Recommended Action: Suggests increasing CPU requests/limits or scaling nodes.

👉 Example Kubernetes CLI Command to View Pod Metrics:

kubectl top pod --sort-by=cpu

🎯 Best Practices for Using DevOps Guru

✅ Tag Resources Effectively: Helps filter insights by application/service.

✅ Integrate with Slack or PagerDuty: Ensures instant alerts for critical issues.

✅ Enable on Business-Critical Workloads: Prioritize production environments.

✅ Review Recommendations Regularly: Improve response time and proactive remediation.

❓ FAQs

Q1: How much does Amazon DevOps Guru cost?

DevOps Guru is a pay-as-you-go service. Costs depend on the number of monitored resources and detected anomalies.

Q2: How long does it take to detect anomalies?

It usually takes a few hours to analyze historical data and detect new issues.

Q3: Can DevOps Guru integrate with third-party tools?

Yes, it supports integrations with Slack, AWS Lambda, PagerDuty, CloudWatch, and SNS for notifications.

Q4: Does DevOps Guru replace traditional monitoring tools?

No, it complements tools like Prometheus, Datadog, or CloudWatch by providing AI-powered insights.

🔥 Conclusion

Amazon DevOps Guru is a powerful AI-driven tool that helps automate anomaly detection, root cause analysis, and proactive remediation. By integrating it into your AWS environment, you can reduce downtime, optimize performance, and improve incident response times.

🚀 Start using DevOps Guru today to bring intelligence to your DevOps workflows and enhance cloud observability!

📢 Have questions or feedback? Drop a comment below or connect with me on Twitter/X@spysood!

Originally published on Medium.