ClusterHQ, arguably the earliest startup to pursue the idea of stateful containers, is calling it quits.

Chairman and CEO Mark Davis announced the shutdown today in a blog entry titled "ClusterF***ed." While he's clearly keeping a sense of humor about things, he's also not taking the decision lightly. "It is disappointing to be at this juncture. It hurts," he writes.

He doesn't explain specifically why the board chose to shut down the company. But he does point out the underlying cause, namely, that "everybody else is talking about stateful containers" nowadays.

ClusterHQ, originally named Hybrid Cluster, launched version 0.1 of its Flocker software in August 2014. Docker Inc. had launched its container runtime earlier the previous year, and the idea of containers had caught on quickly. But being stateless, Docker containers weren't a convenient choice for housing older enterprise applications.

To support a stateful application, a container would need to "remember" certain pieces of data even if the container were to move to a different host. Flocker is an open source volume manager; it's responsible for moving that kind of data alongside the container. Another way to put it is that it lets an operator manage containers and data volumes at the same time.

But as Davis notes, other companies and groups have been piling onto the problem. Some, such as startup Portworx, approach it from a data storage perspective.

Others are orchestration specialists. Mesosphere has declared the era of "Container 2.0," which would include stateful containers. Kubernetes 1.3, released in July, added more support for stateful containers. And Kontena claims it's supported stateful applications all along.

Even Docker Inc. itself has delved into stateful applications, most recently by acquiring startup Infinit.

ClusterHQ raised a $12 million Series A in early 2015.