I have a bunch of blobs stored in Amazon S3. Ruby aws-sdk gem seems to support Swift, we just need to pass endpoint:
swift_client = Aws::S3::Resource.new(
endpoint: 'https://swift.me.com/,
bucket: 'bucket',
credentials: Aws::Credentials.new(access_key_id, secret_access_key))
swift_bucket = swift_client.bucket('bucket')
s3_client = Aws::S3::Resource.new(
region: 'us-east-1',
bucket: 'bucket',
credentials: Aws::Credentials.new(access_key_id, secret_access_key))
s3_bucket = s3_client.bucket('s3-bucket')
# This object exists, s3_object.exists? => true
s3_object = s3_bucket.object('SOMEUID')
swift_object = swift_bucket.object('SOMEUID') # UID is the same as in S3
# I want to copy s3 object to swift, so I do:
swift_object.copy_from(copy_source: "#{s3_bucket.name}/#{s3_object.key}")
At this point I get "Aws::S3::Errors::NoSuchKey: The specified key does not exist.". If I do the same thing vice versa (from Swift to S3), it works fine.
Is it even possible to copy data from S3 to Swift using aws-sdk gem or should I look for another tool?
The solution was in copying content manually, as Swift does not create object at copy_from
execution.
to_object.put(body: object.get.body)