pbs professional 是altair业界领先的用于高性能计算（hpc）环境的负载管理器和作业调度器。
pbs professional 功能
powerful, policy-driven scheduling: pbs professional accelerates job execution and selects optimal job placement across diverse, broadly distributed resources. with pbs pro it’s easy to create intelligent policies to manage distributed, mixed-vendor computing assets as a single, unified system. based on a policy-driven architecture, pbs pro continually optimizes how technical hpc resources are used, ensuring maximized resource utilization and high throughput while respecting business priorities and slas – so more workload is executed faster using fewer resources.
new in 13.0 — more policy controls to better match your business needs
new in 13.0 — expanded scheduling priority formula with full math functions (e.g., sqrt(), ceil(), …), conditional expressions, and a threshold for job start eligibility
new in 13.0 — general fairshare formula enables accruals per-q, license sharing, time-of-day, power use, even combinations of these
new in 13.0 — fine-grained targeting for preemption, configurable at the queue level (admin controlled).
new in 13.0 — million-core scalability: tested to 50,000 nodes, pbs professional scales to support millions of cores with fast job dispatch and minimal latency.
new in 13.0 — fast, reliable startup of huge mpi jobs: pbs professional is tested at tens of thousands of mpi ranks and minimizes delays caused by faulty nodes.
new in 13.0 — fast throughput: pbs professional supports 1,000,000 jobs per day.
new in 13.0 — support for linux control groups: cgroups eliminate resource contention so jobs run faster and don’t interfere with each other or the os.*
new in 13.0 — custom resources can be created directly using qmgr, without the need to restart the server
new in 13.0 — long job and reservation names supported
industry-leading security: pbs professional is the only workload manager to have achieved eal3 certification and the only workload manager offering integration with redhat’s selinux “cross-domain security” (or “mls” — multi-level security) technology*
higher utilization: users can run jobs—or portions of jobs—in the period immediately before a planned outage. typically, computer systems remain unused for several hours prior to outages since insufficient time is available to complete a job before the outage starts. pbs pro fills those holes with malleable “shrink-to-fit” jobs, allowing useful work to be accomplished during the pre-outage period when otherwise no jobs would be running on the system — and thereby providing greater than 95% utilization. jobs get done sooner and scheduling of the computing systems operates more efficiently. one customer reported they were able to reclaim 800,000 cpu hours over two months by using shrink-to-fit to run jobs prior to outages.
flexible plugin framework: pbs professional offers a powerful yet easy to use plugin framework to customize implementations for meeting complex user requirements. for example, ‘execution event’ plugins let users control, modify, extend and change job lifecycle events in the execution stage, allowing for health checks prior to job start, filtering and changing computer behavior when the job starts, and ensuring cleanup is correct.
new in 13.0 — expanded hook events for even more plugin extensibility and customization
green provisioningtm: pbs professional can monitor, shutdown, and restart computing resources based on hpc workload requirements to support enterprise energy conservation initiatives. validated by several leading sites, green provisioning has lowered their energy use by up to 30 percent. .
topology-aware scheduling: pbs professional optimizes task placement for all hpc network topologies (including any topologies built on infinibandc, gige, or proprietary technologies from vendors like cray and sgi), improving application performance and reducing network contention.
cloud computing support: pbs professional is the workload manager for altair’s private cloud appliance . in addition, hpc software-as-a-service (saas) clouds are being powered by the full pbs works suite, including altair's own .
gpu and coprocessor scheduling: pbs professional supports both basic and advanced scheduling to gpus and accelerators (, amd) as well as to the .
enterprise resilience: pbs professional provides a highly-redundant automatic fail-over architecture with no single point of failure — jobs are never lost, and jobs continue to run despite server failures, network failures, and even killing pbs daemons themselves. node health checks can be written to ensure nodes are up and running properly; if a node fails during a job run, the job can be automatically requeued and run elsewhere. custom checks can be added via our extensive plugin framework to ensure continuous operation, even in unique configurations
new in 13.0 — comprehensive health check framework: pbs professional monitors your health check script behavior – either checks run or the node is marked down.
new in 13.0 — expanded support: intel mpi and mpich2 on windows; unc paths for stdin, stdout, and file staging on windows; sles 12 and rhel 7.