No. Even though, NetApp came up with 'flexgroup' to address one-global-file-system concept (by stitching together multiple flexvols), end of the day - "A WAFL filesystem belongs to an aggregate, and an aggregate belongs to one HA pair only", period. To the client it appears as one filesystem. However, internally it's still multiple WAFL volumes. In WAFL flexgroup, a file may be placed entirely on 'Constituent' which is local to the Aggregate. However, a true distributed global-file-system would split the same single file into smaller chunks and spread it across all the Nodes.
One such example is - SpanFS, Cohesity but it's a Backup/Archival/data security product, not a storage appliance YET. However, it is a matter of time.
It is a truly distributed, web-scale file system that exposes industry-standard NFS, SMB, and S3 interfaces across globally distributed environments. So, basically, they already own the backup/archival data layer, so it's only a matter of time before they evolve this into a full-fledged storage appliance and capture the entire end-to-end data foot print . That's my personal view, but I don't think this is far away.
Another one Nutanix, but it is HCI, not truly competing in pure storage space. Another example, WekaFS, again it is not a traditional NAS and lacks all the features that NetApp provides, however, it sits with in the same Storage space, designed for ultra-low latency and parallel workloads.
Comments
Post a Comment