Supportability of Amazon S3 features - Amazon GuardDuty

Supportability of Amazon S3 features

The following table specifies whether or not Malware Protection for S3 supports the listed Amazon S3 features.

S3 feature name Is the support available? Description

S3 Storage Class - S3 Standard

S3 Storage Class - S3 Standard-Infrequent Access

S3 Storage Class - S3 One Zone-Infrequent Access

S3 Storage Class - S3 Glacier Instant Retrieval

Yes

S3 objects can be retrieved without restoring asynchronously.

S3 Storage Class - S3 Intelligent-Tiering

Conditional

  • Intelligent Tiering support is available for S3 objects in the Frequent, Infrequent, and Archive Instance Access tiers.

  • The opt-in Archive and Deep Archive tiers are not supported.

  • Intelligent Tiering always creates a new object in Frequent Access tier. Therefore, object scan on create is supported.

  • Future Intelligent tiering features might start out objects in Archive. Therefore, this is not supported.

S3 Storage Class - S3 Express One Zone (Directory bucket)

No

GuardDuty supports only general purpose buckets for Malware Protection for S3.

S3 Storage Class - S3 Glacier Flexible Retrieval

S3 Storage Class - S3 Glacier Deep Archive

No

The S3 objects must be restored before they can be accessed.

Amazon S3 on Outposts

No

Malware Protection for S3 is not supported on Outposts.

S3 versioning

Yes

All the uploaded S3 objects are scanned for malware. If you uploaded an object with file version v1 and immediately uploaded another version override with v2, then GuardDuty will scan both the object file versions v1 and v2. However, the scan start time might not be in the same order.

S3 Replication - scan replicated object

Yes

If the destination bucket is a protected resource, then GuardDuty will scan all the S3 objects are replicated to the prefixes that are protected and monitored.

S3 Replication: Replicate on scan result tag

No

You can't define a replication rule based on the scan result tag. Amazon S3 does't support replication for tag, except for on create.

Data Encryption - S3-SSE

Data Encryption - SSE-KMS

Data Encryption - DSSE-KMS

AWS KMS - Customer managed key

Yes

GuardDuty supports malware scans for S3 objects that are encrypted with managed and customer managed keys. Ensure that the IAM role includes the permission to use the key. For more information, see Adding IAM policy permissions.

Data Encryption - SSE-C

No

Malware Protection for S3 doesn't support scanning S3 objects that are encrypted with keys that are not accessible.

Client side encryption

No

When your S3 objects are encrypted by using Amazon S3 Encryption Client, your objects aren't exposed to any third party, including AWS. For more information about why this is not supported, see Protecting data by using client-side encryption in the Amazon S3 User Guide.

S3 object lock and legal hold

Yes

Locked S3 objects are locked based on WORM - Write Once Read Many. Malware Protection for S3 can access and scan the objects.

Requester pays

Yes

Malware Protection for S3 can scan the buckets that are set up with Requester Pays. The requester will pay for the S3 calls. For more information, see Using Requester Pays buckets for storage transfers and usage in the Amazon S3 User Guide.

S3: Storage lifecycle

Yes

You can define lifecycle policies based on the scan result tag. For example, auto-delete malicious objects. For more information about lifcycle configuration, see Managing your storage lifecycle in the Amazon S3 User Guide.

S3: Tag-based access control (TBAC)

Yes

You can define bucket resource policies based on your S3 object scan result tag. For example, prevent access to S3 objects that are not yet scanned, or GuardDuty detected threats. For more information, see Using tag-based access control (TBAC) with Malware Protection for S3.