In the request, you must include the computed SHA256 tree hash of the entire archive you have uploaded. For information about computing a SHA256 tree hash, see Computing Checksums. On the server side, Amazon Glacier also constructs the SHA256 tree hash of the assembled archive. If the values match, Amazon Glacier saves the archive to the vault; otherwise, it returns an error, and the operation fails. The ListParts operation returns a list of parts uploaded for a specific multipart upload. It includes checksum information for each uploaded part that can be used to debug a bad checksum issue.
Additionally, Amazon Glacier also checks for any missing content ranges when assembling the archive, if missing content ranges are found, Amazon Glacier returns an error and the operation fails.
Complete Multipart Upload is an idempotent operation. After your first successful complete multipart upload, if you call the operation again within a short period, the operation will succeed and return the same archive ID. This is useful in the event you experience a network issue that causes an aborted connection or receive a 500 server error, in which case you can repeat your Complete Multipart Upload request and get the same archive ID without creating duplicate archives. Note, however, that after the multipart upload completes, you cannot call the List Parts operation and the multipart upload will not appear in List Multipart Uploads response, even if idempotent complete is possible.
An AWS account has full permission to perform all operations (actions). However, AWS Identity and Access Management (IAM) users don't have any permissions by default. You must grant them explicit permission to perform specific actions. For more information, see Access Control Using AWS Identity and Access Management (IAM).
For conceptual information and underlying REST API, see Uploading Large Archives in Parts (Multipart Upload) and Complete Multipart Upload in the Amazon Glacier Developer Guide.