Does copy object to itself causes S3 to realize that it doesn’t need to actually clone the object?

Question

I wish to perform an operation similar to touch(), which just updates the timestamp of the object in s3. However, it seems like there is no such method available in s3. I copy the object to itself to achieve the same purpose. My question is, will copying object to itself be faster than copy the object to somewhere else?

Answer

Please note that as you mentioned correctly, the only way to change the last modified date of an S3 file in the S3 bucket is to copy back the object to the S3 bucket. When an object is copied to the S3 bucket, then in case of versioned buckets, a new object version would be formed while in non-versioned buckets, the older version is replaced by the newer version.

Further, copying objects to the S3 bucket depends on various factors like location of the client making requests, client internet service provider (ISP), client network, client resources and request rate to Amazon S3. But if the locations in both the instances where you are copying the object back to itself or some different location are in the same bucket then there should not be any difference in the speed of uploads.

[+] How do I troubleshoot slow or inconsistent speeds when downloading or uploading to Amazon S3? : https://repost.aws/knowledge-center/s3-troubleshoot-slow-downloads-uploads

Does copy object to itself causes S3 to realize that it doesn’t need to actually clone the object?

Relevant content