What OpenStack Summit 2016 tells us about SDS

Concentric circles in varying shades of green and blue.

Posted on Tuesday, November 8, 2016

Where the annual OpenStack Summit has traditionally been an opportunity for major storage vendors to make loud professions of their support for the technology, the 2016 edition has been markedly different. With the exception of IBM, all of the big storage vendors were conspicuously quiet.

In the previous five summits, Cisco, HPE, EMC and IBM have all maintained a huge presence, trumpeting their commitment to OpenStack. So what happened this year?

The cloud has won ...

The heavy hitters of legacy storage technology have long viewed OpenStack as a way to maintain their place in the corporate data centre. By promoting OpenStack, these vendors had hoped to hold off the Cloud revolution, to keep enterprise customers tied to the traditional on-site data center (and vendor-defined upgrade cycles and maintenance contracts).

But after five years of trying to hold off the inevitable, it seems like everyone but IBM has given up trying to fight the cloud. Or at least their original enthusiasm for OpenStack has cooled significantly.

... or is it a question of revenue?

As the cloud matures, many organizations are discovering problems with the technology. Latency is a major issue for big data deployments for instance, while support for unrestricted storage growth makes cost containment incredibly difficult.

These discoveries should give Cisco and HPE the incentive they need to keep pushing forward with OpenStack. But the problem is not one of uptake, but revenue.

The vendor agnostic nature of software defined storage (SDS) promoted by the OpenStack project means that customers are not tied to storage hardware from a particular vendor. And almost all of the OpenStack deployment case studies delivered during this year’s summit emphasised that fact.

More troubling still (for the OEMs) was that OpenStack adopters are doing so without vendor assistance. In-house teams are liaising with SDS specialists to build their own giant data stores – depriving OEMs of the consultancy fees that may have otherwise helped to cover a shortfall in hardware revenues.

Is this a serious problem?

At the moment it is difficult to say exactly why EMC, HPE and Cisco were virtually anonymous at OpenStack Summit 2016. It could be they are simply re-evaluating their support for the technology, and will return stronger next year. Or it could be a sign that hardware vendors are actively looking for a new alternative to revitalize their flagging storage sales.

There was one undeniable fact to be taken from the summit however – software defined storage is gaining serious traction. Several projects were showcased, outlining how organizations are using SDS to manage vast datasets – including the Square Kilometer Array (the world’s largest telescope) that generates 5,000 petabytes of data every two days.

SDS is in good health – the OEMs perhaps not so much.

To learn more about how SDS can help you break the OEM stranglehold on your data center by redeploying post-warranty storage assets, please get in touch.

Download article as a PDF - What OpenStack Summit 2016 tells us about SDS

More Articles

A mortar board with a cog in the middle, in front of a blue hexagon, and a light blue background

SDS – Making storage more intelligent

SDS storage platforms help to commoditize storage – and make it more intelligent too.

A large red file, and a smaller green file beside it

Four Ways To Save Storage Space

You cannot stop the growth of data, but you can ensure it is stored more efficiently with these four tips.