The NetApp FAS series of storage systems have the potential to be at the center of most enterprise’s storage infrastructure. But there are certain data sets that stretch a FAS too far. For these situations, enterprises need to look for other systems to carry the load. One such use case is unstructured data, the original problem that NetApp tried to solve.
Unstructured data is changing more dramatically than any other data set. While it still contains the original user-created data like documents, presentations and spreadsheets, it also contains potentially petabytes of machine-generated data. Machines generate data much more rapidly than humans do. While it tends to be smaller in size on a per-file basis, machines generate many times more files than their human counterparts, which is breaking the NetApp FAS. The FAS architecture struggles to keep up with the potentially billions of files that machines create and its economics are not ideal for the capacities and retention times this data requires.
The NetApp FAS series is a very versatile family of storage systems. Data centers can leverage FAS systems to host virtual machines on NFS shares and even databases on block storage. NetApp is also still a good storage system for some types of unstructured data, just not all of it. It is best to use a FAS to store unstructured data where high performance matters and where access to that data can come from traditional protocols like NFS and SMB. For near-active and passive data as well as data modern applications access, enterprises should consider object storage.
Object storage can scale to near limitless capacity and file counts. It is also more economically conducive to storing machine data as well as near-active and passive unstructured data. Most importantly, as we discuss in our on-demand webinar “What Does Your Next NetApp Refresh Look Like?“, the ability to move data between NetApp and object storage is improving. New solutions make the access of data between NetApp and object storage, transparent to the users of the data. The result is that IT can drive down the cost of storing unstructured data without impacting the user experience.