Data handling

A method performed by a computing device, the method comprising:

• upon receiving a file corresponding to multiple client-side data units from a client side device,
• identifying a first subset of the multiple client-side data units that are new or revised in comparison with an existing version of the file in a cloud-based datastore and a second subset of the multiple client-side data units that have corresponding copies in one or more second datastore blobs already existing in the cloud-based datastore;
• combining the client-side data units from the first subset of the client-side data units into one or more first datastore blobs, the first datastore blobs individually containing multiple client-side data units from the first subset and having a size that approximates but does not exceed a storage unit size in the cloud-based datastore;
• storing the one or more first datastore blobs in the cloud-based datastore without writing the one or more second datastore blobs in the cloud-based datastore;
• and
• updating a map in the cloud-based datastore to indicate that a new version of the file contains the one or more first datastore blobs written to the cloud-based datastore and the existing one or more second datastore blobs.

A computing device, comprising:

• a processor;
• and
• a memory containing instructions executable by the processor to cause the computing device to:
• in response to receiving, from a client side device, a file corresponding to multiple client-side data units,
• identify a first subset of the multiple client-side data units that are new or revised in comparison with an existing version of the file in a cloud-based datastore and a second subset of the multiple client-side data units that have corresponding copies in one or more second datastore blobs already existing in the cloud-based datastore, the cloud-based datastore storing the existing version of the file in multiple storage units individually having a storage unit size;
• combine the client-side data units from the first subset of the client-side data units into one or more first datastore blobs, the first datastore blobs individually containing multiple client-side data units from the first subset and having a size that approximates but does not exceed the storage unit size;
• store the one or more first datastore blobs in the cloud-based datastore without writing the one or more second datastore blobs in the cloud-based datastore;
• and
• update a map in the cloud-based datastore to indicate that a new version of the file contains the one or more first datastore blobs written to the cloud-based datastore and the existing one or more second datastore blobs.