Object storageΒΆ
Object storage is a very good way to store unstructured data. Unlike a filesystem (where you would have a structured tree of files and folder) or block storage (where you would use a filesystem that kept track of stored blocks), object storage simply stores arbitrary objects.
Each object would typically contain the data that is stored, a varying amount of metadata (that is, information about the stored object) as well as a globally (for the platform) unique ID that reference the object.
Object storage is optimised for storing data that needs to be written once and then read (and re-read). While there is support for committing changes to objects, this is generally avoided. As such, object storage is often used for storing and serving static data, for instance images, PDFs or movies. The backup service in Binero cloud uses our object storage to save the backup data.
In Binero cloud you are able to consume the object storage service using both the S3 API and the Swift API. Both are served from the same underlying storage platform and can reach the same data but use different API formats.
Note
In the documentation we can refer to either a container or bucket both being the same thing, a container holding objects, where bucket is S3 terminology and container is Swift terminology.
You have the ability to chose from several storage policies of object storage to setup a storage solution that delivers at the lowest possible cost per GB while still meeting your requirements. We also enable replication if you want to have your data available in multiple datacenters, for instance as part of a multi datacenter high availability solution.
We recommend reading through our Getting started with object storage guide if you are not previously used to working with object storage in an OpenStack based platform.