Why do Stacks Fail?

• Insufficient permissions (IAM error)
  • Need permissions for all the resources involved!
• Limits exceeded: Too many instances, e.g.
• So many possible reasons

Why would a rollback fail?

• Drift
  • e.g. rolling back to an old (deleted) DB instance
• Not using drift detection
• Trying to delete parent stack before children in nested stacks

Troubleshooting Stack Failures:

• Study the messages during stack creation
• If EC2 issues:
  • Study cloud-init
  • Study logs in /var/log
  • Install cloudWatch logs agent, publish logs
• If one resource is being persnickety:
  • Put it in RetainResources param
  • Delete stack; resource will be retained
  • Delete resource individually
• S3:
  • Non-empty buckets won’t delete