Guidelines and Limitations
The following section provides guidelines and describes limitations when using the AWS Import/Export service. If any of these limitations are important to your use case, contact AWS Support Center with details.
AWS Import/Export does not support Server-Side Encryption (SSE) when importing data.
All objects imported into Amazon S3 have their metadata changed. The only metadata that remains the same is filename and filesize. All other metadata is set as in the following example:
-rw-rw-r-- 1 root root [filesize] Dec 31 1969 [path/filename]
To connect your storage device to one of our AWS Import/Export stations, your device must meet the following requirements:
The device must be compatible with Red Hat Linux and warranted by the manufacturer to support eSATA, USB 3.0, or USB 2.0 interface on Linux.
Maximum device size is 14 inches/35 centimeters high by 19 inches/48 centimeters wide by 36 inches/91 centimeters deep (8Us in a standard 19 inch/48 centimeter rack).
Maximum weight is 50 pounds/22.5 kilograms.
Maximum file size of a single file or object to be imported is 5 TB. Files and objects larger than 5 TB won't be imported. The over-sized files will be listed in your import job log file.
Maximum device capacity is 16 TB for Amazon Simple Storage Service (Amazon S3) and Amazon EBS jobs.
Maximum device capacity is 4 TB for Amazon Glacier jobs.
Power requirements vary by region. For more information, go to the Selecting Your Storage Device section on the AWS Import/Export Product Details page.
You can create up to 50 jobs per day. If you need to create more than 50 jobs, contact your account manager or AWS Support Center.
Ship only one device per package. AWS Import/Export will accept multiple devices in a single shipment only for specific devices, with special packaging requirements. For details, see Shipping Multiple Devices.
AWS Import/Export service does not ship to PO Boxes.
Expedited shipping is available only for return shipments from U.S. data loading facilities to U.S. addresses.
AWS Import/Export exports only the latest version from an Amazon S3 bucket that has versioning turned on.
Amazon DevPay buckets and accounts are not supported. You must use your AWS account credentials and non-DevPay buckets when using AWS Import/Export.
Devices must be healthy. Providing an unhealthy device, such as a device with bad sectors or unstable power, may result in a partial import or export. In extreme cases the import or export will be impossible to perform, and AWS Import/Export might destroy the device. For more information, see Failed Jobs.
Amazon S3 import jobs support the following file systems:
Import to Amazon S3 job file names must be written using UTF-8 character encoding. Any file with a name that is not a valid UTF-8 string is not imported.
You can only perform a Amazon Glacier import from devices of 4 TB in size or smaller. If you need to import more than 4 TB into Amazon Glacier, consider using an Amazon S3 import and archiving your objects to Amazon Glacier with an Amazon S3 lifecycle policy.
Amazon S3 export jobs support the following file systems:
With an Amazon S3 export, some objects might be re-mapped during export to different filenames due to file system limitations. Some of the reasons for remapping filenames include:
The source Amazon S3 object has a non UTF-8 key name.
The Amazon S3 source has two or more object key names that point to the same file name (for instance,
Too many files exist under one directory.
You can find out whether files were remapped for your job by examining the export log uploaded to your log bucket.
If your import job fails, your device will be erased and returned to you. If your export job fails, your device will be returned to you. For more information, see Failed Jobs.