IBM adds a fancy scheduler to its OpenStack stack

IBM modifies its Platform Computing Scheduler to work with OpenStack

IBM has reached into its software portfolio to provide a more sophisticated scheduler for the OpenStack deployments it builds for its customers.

"You don't want a cloud user to feel like their application is not performing as advertised," said Jay Muelhoefer, IBM's strategy director for the new Platform Resource Scheduler.

[ Also on InfoWorld: OpenStack under fire: The critics aren't all wrong. | Is OpenStack the new Linux? Read the early signs around the "cloud operating system." | Track the latest trends in open source with InfoWorld's Open Sources blog and Technology: Open Source newsletter. ]

The Platform Resource Scheduler was originally developed by Platform Computing, which IBM purchased in 2012.

Platform Computing's scheduler, originally called Platform Load Sharing Facility (LSF), found a home in many financial analysis systems, high performance computing (HPC) environments and in a number of grid computing deployments.

Within an OpenStack deployment, the Platform Resource Scheduler can choose the most appropriate server in which to place a new virtual machine based on policies crafted by the administrator. Criteria can include how much of each machine is already being used in terms of its CPU utilization, memory utilization and other factors.

The software will now be included in IBM's SmartCloud Entry and Smart Cloud Orchestrator, two service offerings to help organizations build their own private clouds. It will not be available as a separate stand-alone product, Muelhoefer said.

The scheduler works in conjunction with the OpenStack built-in scheduler, Nova. Nova's limitation is that "it makes decisions based on static information. It only schedules resources once when they are initially placed. It doesn't look how to continuously optimize the environment and revisit the placement decisions," Muelhoefer said.

The software also gives IBM OpenStack users a scheduler that can match the sophistication of VMware's Distributed Resource Scheduler (DRS), found in VMware cloud deployments.

Red Hat has also been busy improving OpenStack's scheduling capabilities. This week, the company released version 4 of its Red Hat Enterprise Linux OpenStack Platform, a package that consists of the latest "Havana" OpenStack distribution, Red Hat Enterprise Linux (RHEL) 6.5, and the Red Hat Enterprise Virtualization Hypervisor (RHEV) . This version comes with the new Heat orchestration engine, and a new resource manager called Foreman.

Joab Jackson covers enterprise software and general technology breaking news for The IDG News Service. Follow Joab on Twitter at @Joab_Jackson. Joab's e-mail address is Joab_Jackson@idg.com

From CIO: 8 Free Online Courses to Grow Your Tech Skills
Join the discussion
Be the first to comment on this article. Our Commenting Policies