Overview
Multipart upload can split an object into multiple parts and upload them to COS. When uploaded, the parts are numbered consecutively. You can upload each part separately or upload them in any order. COS will merge them into an object based on their numbers. If the upload of any part fails, the failed part can be uploaded again without affecting other parts and the content as a whole. In a poor network environment, multipart upload is recommended for objects larger than 20 MB. In a high-bandwidth environment, multipart upload is recommended for objects larger than 100 MB.
Multipart upload can split an object into up to 10,000 parts of 1 MB to 5 GB in size each. The last part can be smaller than 1 MB.
Note:
Simple upload only supports uploading files with a maximum size of 5 GB, while multipart upload supports larger files.
Use Cases
Multipart upload is suitable for uploading large objects in poor or high-bandwidth network environments.
Multipart upload has the following strengths:
In a poor network environment, smaller part size minimizes the impact of restarting a failed upload due to network errors.
In a high-bandwidth environment, multipart upload can maximize the use of your available bandwidth by uploading object parts in parallel. Uploading parts out of order does not affect the final merged object.
With multipart upload, you can pause and resume the upload of a single large object at any time. All incomplete multipart uploads can be resumed unless aborted.
Multipart upload can also be used to begin an upload before you know the final object size. You can initiate an upload and then merge the parts to get the full object size.
How to Use
Using RESTful API
You can use a RESTful API directly to initiate a multipart upload request. For more information, see the following API documents:
Using SDK
You can directly call the multipart upload method in the SDK. For more information, see the SDK documentation for the corresponding programming language below: