IBM Platform Resource Scheduler
Resource-aware management for IBM OpenStack™ clouds
Costly cloud infrastructure that is underutilized and difficult to manage
While OpenStack cloud management solutions are the logical choice for large enterprises and service providers, cloud management can be a challenge. This is especially true as environments grow in size and complexity. Because requirements evolve over time, cloud administrators can find themselves needing to continuously re-balance VM placements to avoid idle resources and over-loaded hypervisors that can lead to performance bottlenecks and application outages.
IBM Platform Resource Scheduler (PRS) helps organizations realize improved service levels and resource utilization in for environments deployed using the IBM Cloud OpenStack Platform IBM smartcloud® environments based on OpenStac, an OpenStack basedk, a cloud management solutionn open-source foundation technology for cloud management.
By extending the basic scheduling capabilities in OpenStack, IBM Platform Resource Scheduler can help reduce the total cost of ownership of public and private clouds by using infrastructure more efficiently, sharing resources between tenants, and automating tedious management tasks. By continuously monitoring the cloud and taking actions as necessary to ensure that configurable policy objectives are being achieved, Platform Resource Scheduler reduces administrative effort and ensures that performance and availability remains optimal even as the topology of the cloud infrastructure and workloads change with time.
IBM Platform Resource Scheduler can be used with other IBM cloud-based offerings including IBM Software Defined Network for Virtual Environments (SDN-VE) and IBM the General Parallel File System (GPFS) used to provide reliable, high-performance distributed storage for OpenStack cloud environments. It can also be used along with IBM Platform Cluster Manager Advanced Edition, a solution that can deploy ready-to-use OpenStack environments along with Platform Resource Scheduler on bare metal infrastructure to simplify provisioning and management.
Better utilization, better service-levels
IBM Platform Resource Scheduler provides
Automated, policy-driven resource management for the IBM Cloud OpenStack Platform
Unlike competing commercial resource managers, or the basic scheduler included in the community OpenStack version, IBM Platform Resource Scheduler provides an open, extensible architecture with rich out-of-the-box policies that gives more control to the Cloud infrastructure teams on how best to share resources to meet a site’s specific requirements. IBM Platform Resource Scheduler is designed to be deployed together with the IBM Cloud OpenStack Platform and value-added offerings such as IBM SmartCloud Orchestrator to help cloud administrators reduce workload while realizing improved utilization and service levels.
Achieve better quality of service
IBM Platform Resource Scheduler helps organizations achieve higher workload quality of service and application performance by avoiding infrastructure bottlenecks and ensuring that VM’s are scheduled to the host best able to meeting unique application requirements.
By exposing a flexible resource selection capability and a rich set of dynamic and static metrics reflecting the state of hosts and hypervisors, cloud administrators and users can make better use of infrastructure resources. Hints can be provided to the OpenStack scheduler in the form of resource requirement expressions and the scheduler will seek to find the optimal resources for particular application workloads to maximize application performance and availability.
Reduce infrastructure and management costs
Using IBM Resource Scheduler, infrastructure teams can reduce costs and save labor by having the scheduler automatically manage large numbers of virtual machines, migrating them transparently and as necessary to meet site specific policy objectives. For example, if a goal is the use the minimum set of resources while still providing adequate services to contain costs, cloud administrators can employ a “packing” policy. For sites that wish to maximize performance, they may wish to employ striping or load-balancing policies, distributing virtual machines widely across multiple hypervisors. The Platform Resource scheduler continuously re-evaluates the state of the cloud and running workloads, automatically making adjustments as necessary to ensure that policies are respected and infrastructure usage is maximized.