Platform LSF

LSF
Developer(s) Platform Computing
Stable release 9.1.2 / February 2013
Operating system Unix, Linux, Windows
Type job scheduler
License Proprietary
Website IBM Platform Computing

Platform Load Sharing Facility (or simply LSF) is a workload management platform, job scheduler, for distributed HPC environments. It can be used to execute batch jobs on networked Unix and Windows systems on many different architectures.[1][2] LSF was based on the Utopia research project at the University of Toronto.[3]

In 2007, Platform released Platform Lava, which is a simplified version of LSF based on an old version of LSF release, licensed under GNU General Public License v2.[4] The project was discontinued in 2011, succeeded by OpenLava.

In Jan 2012, Platform Computing was acquired by IBM.[5]

LSF Scheduling Policies

Fair share, preemptive, backfill and SLA scheduling
High throughput scheduling
Multicluster scheduling
Topology-, resource-, and energy-aware scheduling

[6]

LSF Addon Products

IBM Platform Application Center
Web interfaces for job submission, management and remote visualization.
IBM Platform RTM
A real-time dashboard for monitoring global workloads and resource.
IBM Platform License Scheduler
License management tool with policy-driven allocation and tracking of software licenses.
IBM Platform Analytic 
Analytic tool for visualizing and analyzing workload data.
IBM Platform Process Manager
An interface for designing complex engineering computational processes
IBM Platform Session Scheduler
High-throughput low-latency scheduling solution for LSF environments.
IBM Platform Dynamic Cluster
Cloud management solution to change static cluster into dynamic share cloud resources.

LSF Extensions and integrations

LSF Extensions includes some major extensions:

DRMAA
The Distributed Resource Management Application API handles job management in a range of distributed resource management systems.
HPC Profile Basic
This describes how JSDL, Basic Execution Service (BES) and existing web services security mechanisms can be used interoperable to address batch job scheduling use case.
LSF Perl API
This comprises two modules, Base and Batch, allowing Platform's LSF APIs to be called by Perl.
  • Base module allows Perl applications to link the Load Information Manager (LIM) and Remote Execution Server (RES) daemons for LSF services e.g. retrieving system configuration and dynamic load information for distributed clusters hosts, task placement advice via LIM, and other related functions, thereby improving application performance and resources accessibility.
  • Batch module allows Perl applications to retrieve information as well as the submission of information about the hosts, queues, users, jobs and configuration of the batch system.
SAGA (Simple API for Grid Applications)
The SAGA C++ Reference Implementation provides an LSF plug-in (adaptor) for its standardized job submission, control and monitoring API. The API is available for C++ and Python.
Python LSF wrappers
LSF's API written in C can be easily accessed using Python. Several implementations of LSF Python APIs exist.[7]

LSF is one of the job scheduler mechanisms supported by GRAM (Grid Resource Allocation Manager), a component of the Globus Toolkit.

References

  1. Michael R. Ault, Mike Ault, Madhu Tumma, and Ranko Mosic (2004). Oracle 10g Grid & Real Application Clusters. Rampant TechPress. p. 24. ISBN 9780974435541.
  2. Goering, Richard (March 8, 1999). "Load sharing brings kudos". EE Times Online. Retrieved 2007-11-12. LSF ... enables load sharing by distributing jobs to available CPUs in heterogeneous networks ... but don't tell them that; they'll just want to raise their prices
  3. "Utopia: A Load Sharing Facility for Large, Heterogeneous Distributed Computer Systems". John Wiley & Sons. Retrieved 2007-12-29.
  4. Platform Lava
  5. IBM Closes on Acquisition of Platform Computing
  6. pylsf on GitHub

Further reading

External links

This article is issued from Wikipedia - version of the Wednesday, December 09, 2015. The text is available under the Creative Commons Attribution/Share Alike but additional terms may apply for the media files.