How do I reduce costs for Amazon S3 backups in AWS Backup?

Resolution

Select only required S3 buckets for backups

Complete the following steps:

1. Open the AWS Backup console.
2. In the navigation pane, choose Backup plans.
3. Select the backup plan that's associated with your S3 backup.
4. In the Resource assignments section, select only your required S3 buckets instead of all S3 buckets.
   Note: For more information about resource assignment, see Backup plan resource assignment. If you have an AWS Organizations service control policy (SCP), then change the backup resource selection.

Create a lifecycle rule for the S3 bucket

When AWS Backup backs up a bucket, it backs up the following objects in the bucket:

• Current versions of each object
• Earlier versions of each object
• Delete markers
• Objects that are pending lifecycle actions

Important: There's a minimum charge for each object and marker that you back up. The charge for a delete marker is the same as the charge for a 128 KB object.

To reduce your backup costs, create a lifecycle rule to delete the delete markers and noncurrent versions.

Complete the following steps:

1. Open the Amazon S3 console.
2. In the navigation pane, choose General purpose buckets, and then select the bucket that you want to create a lifecycle rule for.
3. Choose the Management tab.
4. In the Lifecycle rules section, choose Create lifecycle rule.
5. For Lifecycle rule name, enter a name for your rule.
6. For Choose a rule scope, select Limit to scope of this rule using one or more filters.
7. In the Lifecycle rule actions section, select Permanently delete noncurrent versions of objects and Delete expired object delete markers or incomplete multipart uploads.
8. In the Permanently delete noncurrent versions of objects section, for Days after objects become noncurrent, enter the number of days based on your requirements.
   Note: If you don't want to keep noncurrent versions, then keep the Number of newer versions to retain text box empty.
9. In the Delete expired object delete markers or incomplete multipart uploads section, select Delete expired object delete markers and Delete incomplete multipart uploads.

Use continuous backups when the S3 bucket contains large amounts of data

It's a best practice to use continuous backups for large buckets in the following scenarios:

• You back up data at least daily or more frequently, and the data has minimal changes between backups.
• Your bucket has more than 300 million objects.
• You planned your backup lifecycle for more than 35 days. You can set up snapshot backups for the bucket in the same vault where you store your continuous backups.
• You have larger buckets that don't change frequently. You don't need to scan the entire bucket along with multiple requests for each object on preexisting objects.
  Note: Preexisting objects are objects that are unchanged from the previous backup.
• You have buckets that contain more than 100 million objects and have a small delete rate compared with the overall backup size. Your backup plan has a continuous backup with a retention period of 2 days and includes snapshots of a longer retention period.
• Your buckets have both continuous backups and snapshots. AWS Backup takes snapshots from a continuous recovery point so that you don't need to do full bucket scans. For more information, see Compare S3 backup types.

Use the correct storage class for your S3 objects

AWS Backup supports the following S3 storage classes:

• Amazon S3 Standard (S3 Standard)
• Amazon S3 Standard-Infrequent Access (S3 Standard-IA)
• Amazon S3 One Zone-Infrequent Access (S3 One Zone-IA)
• Amazon S3 Glacier Instant Retrieval
• Amazon S3 Intelligent-Tiering (S3 Intelligent-Tiering)

AWS Backup performs multiple calls on each object in an S3 Glacier Instant Retrieval storage class. There's a retrieval cost for these calls. There are also retrieval costs for buckets with objects in S3 Standard--IA and S3 One Zone-IA objects.

To find a storage class that's correct for your environment, see Understanding and managing Amazon S3 storage classes. To determine the costs of S3 storage classes, see the Amazon S3 pricing page.

To identify and change your object's storage class, complete the following steps:

1. Open the Amazon S3 console.
2. In the navigation pane, choose General purpose buckets, and then select your S3 bucket.
3. Select your object.
4. In the Storage class column, identify your object's storage class.
5. On the Actions menu of your object, choose Edit storage class.
6. Select the storage class that you want to change to, and then choose Save changes.

Asses the cost implications for objects that are smaller than 128 KB

Objects in your S3 bucket that are smaller than 128 KB can incur storage costs. The price of S3 objects that are smaller than 128 KB is the same as the price of 128 KB objects.

To identify objects that are smaller than 128 KB, complete the following steps:

1. Open the Amazon S3 console.
2. In the navigation pane, choose General purpose buckets, and then select the S3 bucket that you want to check.
3. Select the object.
4. In the Size column, check for the size of the object.

AWS also charges for GET and List requests on S3 objects, Amazon EventBridge events, and a monthly charge for each gigabyte. For more information, see AWS Backup pricing.

Follow cost reduction best practices

To reduce your S3 backup costs on AWS Backup, take the following actions:

• Use cost allocation tags to monitor S3 backup costs. Then, use AWS Cost Explorer to view and filter your cost allocation tags.
• If you use AWS CloudTrail to log data events or activated server access logging, then deliver logs to a different destination bucket. An infinite loop can cause unexpected charges when you save logs in your source bucket.
• It's a best practice to use S3 lifecycle rules to delete CloudTrail and service access logs based on your time requirements.
• Adjust how you use AWS Key Management Service (AWS KMS), CloudTrail, Amazon CloudWatch, and Amazon GuardDuty as part of your backup strategy.

Note: When you copy a continuous backup across AWS Regions or AWS accounts, the copied recovery point becomes a snapshot backup. Point-in-time recovery (PITR) isn't available for these copies. For more information, see Feature availability by resource and Designing a resilient and cost-effective backup strategy for Amazon S3.