About operations with Amazon Glacier

Last published : Jun 11, 2026
For all the supported cloud types except Glacier, the upload and retrieval operations are synchronous. For Glacier, the retrieval operation is asynchronous and the upload operation is synchronous.
Upload to Glacier
An archive is a base unit of storage, similar to objects in Amazon S3 or Azure BLOB. A vault is a container for storing archives, similar to buckets in S3. An archive is created for each on-premises file, which means that there is a 1:1 mapping of archives and files. When a file is moved to Glacier, all the data in the file is written as one archive in a Glacier vault.
Download or retrieval of an archive from Glacier
Retrieving an archive from Glacier is an asynchronous operation in which you first initiate a job, and then download the output after the job completes. When you initiate a job, Glacier returns a job ID in the response, and executes the job asynchronously. A new fscloudd daemon is added to poll for initiated jobs, and to check whether the job initialization is complete. When a retrieval from the Glacier tier is initiated, the fscloudd daemon is started internally, if it is not running already.
InfoScale supports the following types of retrievals from Glacier:
  • Expedited retrievals, which are typically made available within 1 - 5 minutes.
  • Standard retrievals, which typically complete within 3 - 5 hours. You can access any of your archives within several hours. When a retrieval option is not specified in a request, this option is used by default.
  • Bulk retrievals, which typically complete within 5 - 12 hours. This retrieval option costs the least, and you can use it to retrieve large amounts of data, even petabytes, in a day.
To specify a retrieval type, you provide flags in the policy.xml file.
In the following example, tier2 is a Glacier tier, which indicates relocation from Glacier.
<FROM Flags="expedited_retrieval">
  <SOURCE>
    <CLASS>tier2</CLASS>
  </SOURCE>
</FROM>
Flags for the other supported retrieval types are standard_retrieval and bulk_retrieval.
Relocation from Glacier is done asynchronously, and you can track it by using the fsppadm status command. For example:
\#/opt/VRTS/bin/fsppadm status /mnt1 tier2
UX:vxfs fsppadm: INFO: V-3-20000: fsppadm:
Download Processing for 5 files is remaining.

\#/opt/VRTS/bin/fsppadm status /mnt1 tier2
UX:vxfs fsppadm: INFO: V-3-20000: fsppadm:
Download Processing for this tier is not in progress.