IBM is making a concerted effort to make OpenStack more efficient by delivering tools for simplifying the management of IT that are optimized for the open source cloud-management framework.
A new IBM Platform Resource Scheduler extension to IBM’s SmartCloud Orchestrator leverages technology that was originally developed by Platform Computing to simplify the management of cloud resources. IBM acquired Platform Computing last year.
According to Jay Muelhoefer, director of strategy for IBM Platform Computing, many of the challenges associated with managing cloud resources at scale have already been tackled by Platform Computing, which developed a platform that was specifically optimized to manage high-performance computing (HPC) environments. After making a major commitment to OpenStack earlier this year, Muelhoefer says that IBM is now beginning to layer on top of OpenStack a suite of management technologies that promise to make it easier to dynamically manage cloud computing at scale as part of a larger software-defined environment (SDE) that IBM is building on top of OpenStack technologies.
Managing IT at scale represents the single biggest technical challenge facing organizations as they embrace cloud computing. Most organizations can’t afford to hire an endless number of administrators, which means relying on more automation technologies, such as IBM Platform Resource Scheduler, to dynamically allocate resources on demand across thousands of virtual and physical servers.
Having already faced many of the same issues in HPC environments, Muelhoefer says many of the IT automation technologies originally designed to manage scientific applications running on HPC systems can now be applied to cloud computing. In order back up its commitment to OpenStack, IBM is now closely coupling such tools to the rest of its OpenStack software portfolio.
Muelhoefer says one of the specific issues that IBM is trying to address with IBM Platform Resource Scheduler is that in its current form, OpenStack lacks a dynamic ability to reallocate resources. Each instance of OpenStack creates a static distribution of workloads that can’t be reallocated without rebooting the entire implementation. In time, the companies that make up the OpenStack consortium are expected to address this issue in future updates. But for now, Muelhoefer says IBM is tying IBM Platform Resource Scheduler to its IBM SmartCloud Orchestration software for OpenStack environments to address the issue. In the near future, IBM also plans to make IBM Platform Resource Scheduler an element of its IBM SmartCloud Entry self-service portal offering.
As a platform for managing enterprise IT, the OpenStack platform remains a work in progress. But as time goes on, it’s clear that OpenStack has a lot to offer as an extensible framework for managing cloud computing deployments based on policies set by the internal IT organization. Ultimately, the battle for supremacy in the cloud will come down to which cloud management framework not only costs less to implement, but also can be trusted most by the IT organization.
While OpenStack clearly still has a long way to go before it can rival proprietary management platforms from vendors such as VMware, Microsoft and Citrix in terms of completeness, the one thing it does have going for it is its ability to provide a framework that is highly customizable, and a lot more transparent in terms of what the algorithms being employed are actually doing to optimize the cloud computing environment.