Why are object stores a good match for archival data storage?

Object storage is a software layer that sits on a commodity x86 server, which is why many vendors call it software-defined storage. The technology takes advantage of embedded storage inside that server, creates a node and then interlinks multiple nodes to create an object storage system across Ethernet.

Object stores, by definition, store data as objects. Even though it's writing to media, such as a hard disk drive or even flash -- in the form of blocks -- it interfaces to applications across TCP/IP using HTTP puts and gets, which are referred to as REST or a RESTful interface. Object storage adds a lot of metadata to every object. In fact, you have a lot of flexibility as to how you describe each object and the metadata stored with the data. This is in comparison to a file system, for example, where metadata is separated from the data and usually put in a database.

Due to that extra layer of processing, you have more metadata and process data between the application and the actual media you are writing to, which means you're adding latency. And because object storage tends to be slower than other types of storage, object stores are generally best suited for archival data storage, backup data and secondary data applications -- not frequently accessed applications or apps that have any kind of interactive or transactional requirement.

You could use object stores to replace large, secondary storage arrays and file-based storage arrays. Additionally, you can replace tape libraries, because object storage makes a highly resilient, highly durable, active archive and even a passive cold archive.