This operation downloads the output of the job you initiated using InitiateJob. Depending on the job type you specified when you initiated the job, the output will be either the content of an archive or a vault inventory.

A job ID will not expire for at least 24 hours after Amazon Glacier completes the job. That is, you can download the job output within the 24 hours period after Amazon Glacier completes the job.

If the job output is large, then you can use the Range request header to retrieve a portion of the output. This allows you to download the entire output in smaller chunks of bytes. For example, suppose you have 1 GB of job output you want to download and you decide to download 128 MB chunks of data at a time, which is a total of eight Get Job Output requests. You use the following process to download the job output:

Download a 128 MB chunk of output by specifying the appropriate byte range

using the Range header.

Along with the data, the response includes a SHA256 tree hash of the payload.

You compute the checksum of the payload on the client and compare it with the checksum you received in the response to ensure you received all the expected data.

Repeat steps 1 and 2 for all the eight 128 MB chunks of output data, each

time specifying the appropriate byte range.

After downloading all the parts of the job output, you have a list of

eight checksum values. Compute the tree hash of these values to find the checksum of the entire output. Using the DescribeJob API, obtain job information of the job that provided you the output. The response includes the checksum of the entire archive stored in Amazon Glacier. You compare this value with the checksum you computed to ensure you have downloaded the entire archive content with no errors.

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 the underlying REST API, go to Downloading a Vault Inventory (, Downloading an Archive (, and Get Job Output (