Hey there, I’ve run into similar head-scratching issues with S3 lifecycle rules not cleaning up expired objects as expected, so let’s walk through the most common checks and fixes to get to the bottom of this:

• Verify Your Lifecycle Rule Configuration
  First, double-check that your rule is actually set up correctly and targeting the right objects:

  • Use the AWS CLI command aws s3api get-bucket-lifecycle-configuration --bucket your-bucket-name to pull the exact rule details. Look for typos in expiration dates, or if you used DaysAfterCreation instead of a specific date, confirm the timeline adds up to your 3-month expiration window.
  • Make sure the rule is applied to the correct bucket—it’s easy to accidentally configure the rule on a different bucket if you’re managing multiple.
  • Check if the rule has any filters (like prefixes or tags) that might be excluding your expired objects. For example, if your rule only applies to objects with a cleanup: required tag, but your 3-month-old objects don’t have that tag, they won’t be deleted.

• Check for Versioning or Object-Specific Expiration Settings
  If your bucket has versioning enabled, standard expiration rules only handle the current version of objects. Old versions (noncurrent versions) require a separate NoncurrentVersionExpiration rule to be cleaned up—this is one of the most common gotchas!
  Also, verify if any of your expired objects have an individual x-amz-expiration header set. This per-object setting takes priority over bucket-level lifecycle rules; if the header’s date is set to something later than March, the object won’t be deleted yet.

• Dig Into Logs for Clues
  S3 logs and CloudTrail can tell you exactly what’s happening (or not happening):

  • Enable bucket access logging if you haven’t already, then look for entries related to lifecycle expiration. You might see failed deletion attempts with error codes that point to issues like permission problems or object locks.
  • Use CloudTrail to search for LifecycleExpiration events. Filter by your bucket name and check if there are any failed events—this will show you why S3 couldn’t delete the objects (e.g., missing IAM permissions for the lifecycle service, legal holds on objects, etc.).

• Rule Out Edge Cases
  While S3 typically runs lifecycle rules daily, there are rare scenarios that cause unusual delays:

  • If your bucket has millions of objects, lifecycle processing might take longer than usual, but a multi-month delay is way outside the norm. If this is your case, reaching out to AWS Support can help confirm if there’s a backlog affecting your bucket.
  • Double-check when you actually created the lifecycle rule. If you only set it up recently (not back in March), that would explain why objects haven’t been deleted yet.

If none of these steps resolve the issue, your best bet is to contact AWS Support directly. Provide them with your bucket name, the lifecycle rule details, and any relevant log entries—they can access internal metrics to figure out why the expiration isn’t triggering.

内容的提问来源于stack exchange，提问作者Alex Levine