This is one of a series of posts discussing the presenters at Storage Field Day 10, occurring 25th-27th May 2016 in Silicon Valley, at which I am an attending delegate. See links at the end of this post to other presenting vendors. Details of the event can be found on the Storage Field Day 10 page at techfieldday.com and on the dedicated events page on this site at Storage Field Day 10.
There have been many attempts over the years to abstract the management from the access of data in shared storage environments. Some operate inline, like IBM’s SVC and HDS’ Universal Volume Manager. Others operate out-of-band and provide point-to-point connectivity while separating the metadata that manages the association of storage sources and targets.
Primary Data is the second of these solution types, connecting storage “providers” and “consumers” primarily using NAS protocols such as NFS and SMB in a product solution known as DataSphere. Out-of-band management is provided through a redundant pair of metadata controllers that track the associations between source and target, implementing the ability to snapshot and replicate data.
As a concept, data abstraction or virtualisation (however you want to define it) is a good idea. It provides mobility (up to the extent of the virtualisation layer itself), allows the reuse of old storage resources and in the NAS world, implements a single consistent name space.
However, introducing an abstraction layer also introduces complexity. If data for a single entity (file, object or volume) is spread across multiple hardware platforms, then the resiliency of the solution is based on the lowest common denominator – the system with the lowest availability. Implementing data protection at the abstraction layer makes little sense, as it duplicates a function that lower levels of hardware are already doing. It also wastes resources. Bear in mind that other distributed solutions will replicate data but have no redundancy within a single device.
As a result, I find it difficult to see value in the Primary Data solution, outside the benefit of providing some mobility features. The reason I think this, is because many of the storage management issues in the data centre today could have been solved though better practices, like implementing good standards and processes. Admittedly, NAS data has specific challenges, not least of which is how to create a logical name space that isn’t constrained by physical appliances. There are however, plenty of other solutions that can solve that problem. Implementing abstraction and mobility for data on highly resilient systems doesn’t seem to make sense, when there are other solutions around which can distribute and abstract the data on commodity hardware without having to buy and implement a relatively expensive appliance.
Perhaps things have evolved with DataSphere and there are some new features to talk about. I’m still on the fence on this one and need to be convinced there is real value in this solution.
Primary Data has been a previous Storage Field Day attendee. Links to their presentations can be found below.
- Primary Data Presents at Storage Field Day 7 (Tech Field Day website)
- Primary Data Presents at Storage Field Day 8 (Tech Field Day website)
Comments are always welcome; please read our Comments Policy first. If you have any related links of interest, please feel free to add them as a comment for consideration.
Disclaimer: I was personally invited to attend Storage Field Day 10, with the event teams covering my travel and accommodation costs. However I was not compensated for my time. I am not required to blog on any content; blog posts are not edited or reviewed by the presenters or the respective companies prior to publication.
Copyright (c) 2009-2016 – Chris M Evans, first published on https://blog.architecting.it, do not reproduce without permission.