GigaSpaces, a provider of an in-memory computing platform, has added support for a Kubernetes Operator that makes it simpler to deploy and update its platform.
Lee Blum, director of product management for GigaSpaces, said the company is making available a Kubernetes Operator based on the management framework originally defined by CoreOS, alongside existing support for the open source Helm tool, simplifying deployment and management of its platform on a Kubernetes cluster.
While there are still many organizations employing Helm, Blum says it’s becoming apparent most organizations are starting to rely more on Operators to manage their Kubernetes environments.
The Kubernetes Operator has been made available as part of release 15.8 of the GigaSpaces InsightEdge Portfolio, which reduces the memory footprint for the platform while simultaneously improving SQL query and business intelligence (BI) performance.
In general, Blum says there is a high correlation between adoption of Kubernetes and its own in-memory computing platform. As organizations shift toward building applications that run in near-real-time to drive digital business transformation initiatives, there’s a natural tendency to also employ a modern platform designed to make it simpler to scale compute resources up and down, as needed.
While the scope of a digital business transformation will vary by organization, the one thing they tend to have in common is an event-driven IT platform that drives real-time processing. IT organizations are starting to process and analyze data in near-real-time at the points where it is created and consumed. Applications based on batch-oriented processing have been in place for decades, and were not designed to provide the level of interactivity required by a digital business process. Batch-oriented applications are typically updated overnight, which means the data they contain can be out of sync with the true state of the business by as much as 24 hours.
There’s no shortage of in-memory computing platforms, which will play a critical role in driving real-time applications. GigaSpaces is making the case for a platform that is both more accessible via application programming interfaces (APIs) and also is cost efficient. With the latest release, RAM usage has been optimized to reduce infrastructure costs by up to 70%, says Blum.
At the same time, a smart data locality capability added to the latest release boosts SQL query performance by a factor of ten. The capability automatically replicates selected small data tables to the nodes in the cluster to accelerate server-side JOIN performance. A Broadcast Objects feature also replicates smaller, static tables that are used frequently when combining rows from two or more tables.
It’s not clear at what rate in-memory computing platforms will replace traditional database platforms that access data stored on hard drives. What is clear is the bulk of new applications tend to be very sensitive to latency. As the cost of memory continues to decline, more applications that run in memory will be deployed. In fact, running applications in memory may soon become the de facto standard.
