Amazon Glacier Backup

Since version 6.20, Syncovery can upload files to Amazon Glacier. To set up Glacier as the destination, click on the Internet button for the right-hand side and choose Glacier as the Internet Protocol.

Amazon Glacier is intended for long term archiving and backup with the expectation that most operations are uploads and not downloads. Downloads should be a rare exception, since they can be extremely slow. To download a file, Syncovery needs to initiate a job with Amazon Glacier, and even with small files, it takes at least three hours for Amazon Glacier to retrieve the item. The same is true for file listings. Therefore, Glacier cannot be used for two-way synchronizations. It is intended only for backing up and archiving.

Please be aware that there is a download fee from Amazon if you download more than 5% of your data per month (or something like that). It seems that in some special cases, the download fee can be extreme. If you intend to make a full backup and restore for testing, please read the fee schedule closely, and also read articles such as this one:
Is There a Landmine Hidden in Amazon’s Glacier?


Two methods to use Glacier

  • Choose S3 inside our software, and use a lifecycle rule in S3 to move files to Glacier to reduce costs. See the Amazon Web Services Blog
  • Choose Glacier directly in our software. The rest of this article describes this method.

Because there is no quick way to get a file listing from Glacier, Syncovery maintains a local database with the information of all the files on Glacier. This database is automatically maintained and it is independent from the checkmark “Cache Destination File List”. Therefore, with Glacier, there is no need to use the caching option. The local databases used for Glacier are separate for each vault, but shared among all profiles. When you create a new profile that accesses an existing vault, it will alraedy know the vault contents, if the vault has been used on the same machine.

Glacier also cannot rename items which have been uploaded. When you rename or move files on the local side, they have to be uploaded again.

In version 6.20, the download process is still inefficient if you need to download many files. It will initiate the download job for each file and wait for it to complete, then download the file, and then proceed to the next file. You can specify a number of files to copy in parallel in this fashion (on the Files tab sheet), but the limit is 30. So, you can download up to 30 files in 3 to 5 hours, but downloading 300 files will take 30 to 50 hours. This limitation will be removed in a future update.

In light of this, if there is a chance you will need a complete restore, it may make sense to use “Zip Packages” with many files per zip archive, instead of uploading individual files.

This information will be updated as support for Glacier is optimized and more experiences are made with Amazon’s new service.