DevOps Tools and Cloud Platforms – Service Limits, Quotas, and Best Practices

This document provides an overview of common service limits and quotas for major cloud platforms (AWS, Azure) and other DevOps tools (Terraform, Docker, Kubernetes, Git, Jenkins, etc.), as well as some basic troubleshooting scenarios and best practices.

---

AWS Service Limits and Quotas

1. EC2 (Elastic Compute Cloud)

• VMs per Region: 20 (default)
• Maximum EBS Volumes: 1,000 per Region
• Maximum Instance Types per Region: Varies (request increase)

2. VPC (Virtual Private Cloud)

• Maximum VPCs per Region: 5
• Subnets per VPC: 200
• Security Groups per VPC: 500

3. S3 (Simple Storage Service)

• Object Size: 5 TB
• Bucket Name Length: 63 characters
• Maximum Buckets per Account: 100

4. IAM (Identity and Access Management)

• Users per Account: 5,000
• Groups per Account: 300
• Roles per Account: 1,000

5. RDS (Relational Database Service)

• DB Instances per Region: 40
• Maximum DB Storage: 64 TB

6. SQS (Simple Queue Service)

• Message Size: 256 KB
• Maximum Queue Length: Unlimited

7. Lambda

• Max Concurrent Executions: 1,000
• Memory per Function: 10 GB
• Execution Timeout: 15 minutes

---

Azure Service Limits and Quotas

1. Virtual Machines (VM)

• VMs per Subscription: 20,000 (default)
• Maximum Disk Size: 4 TB
• Maximum VM Network Interfaces: 8 per VM

2. Virtual Network (VNet)

• VNets per Region: 50
• Maximum Subnets per VNet: 1,000
• Maximum Public IPs per Subscription: 1,000

3. Storage Accounts

• Maximum Storage Accounts per Subscription: 200
• Maximum Blob Size: 5 TB
• Maximum Containers per Account: 500,000

4. Kubernetes Service (AKS)

• Nodes per Cluster: 5,000
• Pods per Cluster: 150,000
• Node Pools per Cluster: 100

5. Azure SQL Database

• Databases per Server: 500
• Maximum Storage Size: 16 TB
• DTUs per Database: 4,000

6. App Service (Web Apps)

• Web Apps per Subscription: 200
• Max Instances per Plan: 20,000

---

DevOps Tools

1. Terraform

• Loops: count, for_each
• Conditional Logic: Use count and if statements
• Local Exec Provisioner: local-exec for running commands
• Example: Using count to create multiple resources

  resource "aws_s3_bucket" "buckets" {
    count = 5
    bucket = "my-bucket-${count.index}"
  }

2. Docker

• CMD vs ENTRYPOINT:
  • CMD: Provides default arguments for the container. Can be overridden at runtime.
  • ENTRYPOINT: Defines the default executable. Cannot be overridden.
  • Example:

    # CMD example
    CMD ["python", "app.py"]
    # ENTRYPOINT example
    ENTRYPOINT ["python"]

3. Kubernetes

• ConfigMaps: Store configuration data in key-value pairs.
• Pods: The smallest deployable unit.
• Services: Expose pods to access from outside the cluster.

4. Git

• Fetch vs Pull:
  • Fetch: Downloads changes but does not merge.
  • Pull: Fetches changes and merges them into the local branch.
• Cherry-pick: Apply a commit from another branch.

  git cherry-pick <commit-hash>

• Amend: Modify the last commit.

  git commit --amend

---

Scenario-Based Troubleshooting

1. AWS EC2 Instance Unreachable

• Issue: EC2 instance cannot be reached.
• Fix:
  • Check Security Group rules.
  • Ensure NACL is not blocking the traffic.
  • Verify the Route Table configurations.

2. Azure VM Not Starting

• Issue: VM stuck in a "starting" state.
• Fix:
  • Check the boot diagnostics for error logs.
  • Ensure there are no issues with VM Size availability.
  • Verify that the Storage Account used by the VM is not full.

---

Best Practices

1. Security

• Use MFA and IAM Roles for secure access.
• Encrypt sensitive data in services like S3, RDS, and Azure Storage.

2. Backup and Disaster Recovery

• Cross-region replication for AWS and Azure to ensure disaster recovery.
• Regularly test backup and restore procedures.

3. Scaling

• Use Auto Scaling for EC2, Azure VM Scale Sets, and Kubernetes to handle varying loads.

---