Hammerspace, a provider of storage software, today released the results of a survey that finds when it comes to deploying stateful applications on Kubernetes clusters, most organizations are still encountering significant challenges.
Based on 219 respondents of a survey conducted during the recent online KubeCon + CloudNativeCon North America 2020 conference, the results find 80% of the attendees are challenged by the time and effort required to get stateful applications running on Kubernetes clusters. Most of the respondents said the process is not as automated as it could be, which inhibits the ability of DevOps teams to self-service their own requirements. More than two-thirds of respondents (67%) want to simplify data orchestration to enable DevOps self-service.
Brendan Wolfe, vice president of product marketing for Hammerspace, says that complexity is inhibiting the deployment of stateful applications on Kubernetes clusters that still largely run stateless applications. Most of those stateless applications eventually store data on a storage system that is external to the Kubernetes cluster. The issue that creates is many IT teams would prefer to reduce costs by centralizing the management of compute and storage on the same Kubernetes cluster.
Overall, the survey finds the top factor when selection Kubernetes storage technologies is complexity (40%), followed by cost (30%). The fact that complexity is still a larger concern than cost speaks to the extent of the challenge IT teams currently face, says Wolfe.
Hammerspace is making the case for a software-defined approach that unifies the management of data and storage on a Kubernetes cluster. Designed to run natively on a Kubernetes cluster, that approach makes it feasible to orchestrate hundreds of thousands of stateful containers that are being updated constantly, says Wolfe. In fact, as stateful container applications are increasingly deployed in production environments, Wolfe notes most IT organizations have yet to come to terms with the need to finally unify storage and data management in those environments. Most IT organizations currently isolate storage data management within legacy IT environments, resulting in silos that only serve to increase the total cost of IT, he notes.
It’s not clear going forward to what degree data operations and DevOps might actually converge within IT environments as stateful applications become more widely deployed. A recent survey conducted by the Cloud Native Computing Foundation (CNCF) that oversees the development of Kubernetes found 55% of respondents have now deployed stateful container applications in production, with another 11% planning to deploy them in the next 12 months. Another 12% are evaluating them, according to the survey. As such, it’s now only a matter of time before organizations at the very least revalue the role storage administrators might play inside their organization.
Of course, not every organization can afford to have a dedicated storage administrator. In many cases, a DevOps team has already subsumed the provisioning and management of storage resources within its workflows. Regardless of who manages data and storage, it’s apparent legacy approaches to managing those systems are not going to be quite the same as it once was.