OpenStack – A cloud of dreams

OpenStack is an Open Source cloud platform, created in 2010. Despite high-profile supporters it faces a number of questions about its future in the highly competitive cloud market.

The CEO of Rackspace (creator of OpenStack) said in December that the main market is Fortune 500 companies building private clouds. Rackspace has recently pivoted to offer support for AWS and Azure. HPE, another major OpenStack supporter, has also pivoted with Azure now its preferred public cloud partner. HPE’s Cloud28+ initiative has moved away from being 100% OpenStack to supporting Microsoft and VMware. OpenStack seems to be struggling and so what are the issues?

Is it really cost effective?

OSS is often considered the lower cost option. Only if you’re prepared to put in the hard graft yourself and not rely on vendors that make it easy. Linux is free but organisations pay for distributions with support. As a niche player, OpenStack skills are hard to find and therefore can be expensive. Such is the shortage of skills that some OpenStack vendors and SIs are offering them to customers as premium services. That some predict the OpenStack market to be worth over $3 billion by 2018, when the product is free, should be a giveaway that all is not what it seems.

Lacks public cloud scale

Public cloud is dominated by rich vendors investing in global infrastructures that make consumption highly cost-effective for the customer. Open Stack lacks such support. This leaves customers who wish to deploy across multiple geographies having to negotiate multiple commercial relationships with different providers. This not only adds complexity and possible cost but also leaves them far more at the mercy of commercial forces. There may be good reasons for some to choose this route in the short term such as very specific localised requirements. But in the longer term this approach should be less appealing.

Private cloud market will shrink

Even if one believes that organisations will not move everything to the public cloud (a sensible belief) there is no doubt that many are reducing their infrastructure in favour of public clouds and this will continue. OpenStack would be selling into a shrinking market. Also, if organisation plan to one day move to public cloud why would they invest in a platform that has no major public cloud support? With the potential of having to re-architect and re-engineer solutions.

Interoperability is more dream than reality

The ability to move a workload from one cloud to another without having to make changes, but is it sensible? The true value of cloud is in embracing cloud native services. Simply moving an on premise Oracle DB to a VM in the cloud will derive minimal benefit compared to using a cloud native data store such as Amazon S3. Also, different distributions of OpenStack will break interoperability and that’s already been happening. OpenStack has tried to stop this by getting vendors to agree common standards, but if they do that then how will they compete? By offering custom services that if used will negate interoperability.

For a few, not the many

There will of course be organisations for whom OpenStack is appealing such as Fortune 500. They will continue to run their own large data centres for some time and be looking to harness the power of cloud with some very specific needs that OpenStack may meet. But these types of organisation will be few and shrink over time as they too move to public cloud. Where does this leave OpenStack over the long term?