Hello all, I started doing some large COPY from box.com to amazon as a backup, seems to work fine going right to glacier, as with --size-only the metadata of an object has the info we need to compare
just wanted to confirm, it appears that the exposed AWS api does not allow for glacier objects to be moved / copied to a different bucket or folder.
I am wondering if anybody has come up with a sync methodology using glacier to move files to a --backup-dir
thanks
Does server side copy not work on GLACIER objects?
I suspect this is the relevant paragraph in the docs:
If the source object is archived in Amazon Glacier (the storage class of the object is
GLACIER), you must first restore a temporary copy before you can copy the object to another bucket. For information about archiving objects, see Transitioning to the GLACIER Storage Class (Object Archival).
So looks like you have to restore the files then you can copy them.
s3 doesn’t support move only COPY then DELETE which is what rclone does when you
Is there a feature request to add this to the s3 support? 
Might be ugly since restore takes time - I guess generate a list of what would have gone to backup dir, restore it, copy, delete, sounds ugly
I don’t think so…
Though there is the settier command which would be the right place to plumb in support.
Yes, rclone would need a list to restore…
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.