June 24, 2017

Red Hat is starting to make a case for coupling its container platform with storage systems running its Gluster software.

Following the general availability of version 3.2 of Gluster being made today, Sayan Saha, head of product for Red Hat Gluster Storage, says organizations building stateful applications using containers can now take advantage of storage services such as geo-replication and in-flight encryption. Delivered as a Docker image, those services will make it possible to run three times more persistent volumes per cluster, Saha says.

In addition, improvements to how Gluster’s network attached storage system processes metadata can improve processing of requests for files by container registries by a factor of eight, he says.

When coupled to the Red Hat OpenShift Container Platform, Saha says IT organizations now can rely on a single vendor to provide the software infrastructure needed to support stateful applications based on containers. Most developers, he says, prefer to access storage resources via a Red Hat OpenShift Container Platform because they don’t want to directly access underlying storage resources. They want to be able to access data without the storage system getting in the way.

Other new capabilities found in Gluster 3.2 include faster self-healing capabilities and an arbiter volume feature that resolves conflicts stemming from data mismatches that might occur. That’s significant, Saha notes, because it provides the benefits of three-way replication with consuming additional infrastructure resources.

Finally, Red Hat is adding support for asynchronous notifications via the Nagios open-source monitoring framework that Red Hat provides for Gluster storage deployments.

As IT organizations start to build and deploy more robust applications using containers, the way microservices interact with underlying storage systems becomes critical because there’s more potential for contention. Where once there were just a few dozen virtual machines trying to access data, it’s now possible to have a scenario where hundreds of microservices are trying to access a common pool of data. In many cases, the deployment of those microservices will force IT organizations to upgrade their storage systems to ensure consistent levels of performance. It still relatively early days when it comes to microservices, but Saha notes that many storage administrators are coming up to speed rapidly on the unique attributes of microservices.

Naturally, competition is starting to intensify among storage vendors looking to add support for classes of workloads that usually need access to high-performance storage systems. Most of them already support Docker containers natively, and it’s now only a matter of time before most of them natively support container orchestration engines such as Kubernetes. The concern for most providers of commercial storage systems is to what degree the shift to containers will pull IT organizations toward open-source storage systems. Red Hat is betting that’s precisely what will happen.

In the meantime, most IT organizations are still coming to terms with containers and microservices. In the absence of any other management tools, most containers are still being deployed on top of virtual machines. But as the number of stateful container applications increase, interest in running containers on bare-metal servers to improve overall I/O throughput also will increase.

Mike Vizard

Mike Vizard is a seasoned IT journalist with over 25 years of experience. He also contributed to IT Business Edge, Channel Insider, Baseline and a variety of other IT titles. Previously, Vizard was the editorial director for Ziff-Davis Enterprise as well as Editor-in-Chief for CRN and InfoWorld.