For large scale cloud computing, object storage is the best. Object storage is best for solving the increasing problems of data growth as more data is being generated, storage systems have to grow at the same pace.
In object storage you do not have the ability to edit one part of a file, all objects have to be manipulated as a whole unit. Object data can scale infinitely to petabytes, data can be stored across many regions.
Object data also allow the use of customized metadata tag for better identification, classification and retriving of data.
There are two main kinds of platform for large scale cloud storage: Block storage and Object storage. Which is best depends upon what you are storing. It has been found that standard IP technology based in TCP does not work well for large-scale uploads to cloud storage. Block storage provides a virtualized Storage Area Network with logical volume management provisioning via a simplified web services interface.
It must be provisioned prior to use, but it's something that can be configured to operate a specific number of input/output operations per second; ideal for software that interacts directly with disk drives. Object storage is better for storing large amounts of unstructured data. This is a flexible and cheap option. Use it for data-intensive cases such as cloud-based media processing, big data analytics, and long-term archiving and backup.