Monday Seminars
Grid Engine Admin Training — Chris Dagdigian, Daniel Templeton
This day-long seminar is intended to provide a solid overview of Grid Engine administration in a lecture hall setting. This is not an advanced"" administration workshop as Grid Engine sourcecode details, internal data structures and the DRMAA API will not be typically addressed.
The material being presented is normally part of an interactive classroom program with live demonstrations and training clusters, we are modifying the original contents to best suit the lecture hall setting in Regensberg.
Topics to be covered include:
• Installation & Deployment Considerations
• CLI, scripted & graphical installation options
• CLI Configuration
• SGE 5.x vs. 6.x vs. 6.1x vs. 6.2x
• SGE Usage for Admins
• Monitoring & Reporting
• Policies & Resource Allocation
• Resource Quotas
• Application & Workflow Integration
• ARCo & SGEInspect
• Troubleshooting
Time permitting we may also cover service domain management (SDM) and the cloud services adaptor.
HPC Application Development — Ruud van der Pas
An HPC Parallel Programming Seminar will teach aspects of parallel computing as well as providing a good understanding of the basics of the MPI and OpenMP programming models. Some basic understanding of C or Fortran will be helpful, but it is not a requirement.
Lustre Advanced Topics Seminar — Johann Lombardi, Andreas Dilger
A special Lustre Advanced Technical Seminar is offered with senior Lustre engineers. They will discuss areas of the product they are working in and take questions and provide answers around tuning, configuration, and debugging.
Workshop Sessions
Building Grid Engine Clusters in Amazon EC2 — Chris Dagdigian
Chris will discuss practical tips and tricks for people interested in deploying Grid Engine on the Amazon EC2 compute cloud. We will also address the question of why would anyone want to do such a thing?
Cloud Computing at Sun — Roland Rambau
In this session, we'll have a look at some of the Cloud Computing related projects happening at Sun. We'll discuss the Sun Cloud, developing applications for the Cloud with project Kenai and Zembly, as well as ready-to-go Cloud toolkits such as the Immutable Service Containers project or pre-installed OpenSolaris AMIs.
Clusters, Clouds, and Service Domain Manager — Torsten Blix, Michael Bachorik
This session will introduce the new Service Domain Manager component which is part of Grid Engine. It will explain underlying concepts like 'Resources', 'Services' and 'Service Level Objectives'. It will be shown how SDM can be used to implement elastic multi clustering for Grid Engine, including the utilization of external cloud resources. A live demonstration will show how SDM could be used to manage Solaris Zones as cluster resources.
Data Management with Shared QFS and Storage Archive Manager (SAM) — Harriett Coverston
Learn about Sun QFS shared SAN file system, the only distributed file system offered by Sun with tightly integrated data management based on the innovative Storage Archive Manager (SAM) technology. If you are struggling with backup, learn how SAM changes the paradigm of traditional backup. Learn how time-based archiving can automatically and continuously protect data with SAM policy management and dynamic placement of data across tiers of storage saving you money. If you application requires streaming I/O learn how QFS can deliver data at near raw speeds or distribute content in web based applicaitons. If you are a Lustre user, learn how SAM-QFS provides data protection for Sun's HPC Solutions. Learn about all the high availability features in our 5.0 release and learn about our improved metadata performance.
Data-Aware Routing in the Cloud — Eugene Steinberg
Large-scale HPC infrastructures such as Sun Grid Engine sooner or later face data access issues, usually caused by the centralized nature of data sources. This talk discusses an approach to mitigate those issues by employing in-memory data grids, such as Oracle Coherence. By collocating compute and data grid nodes on the same hosts, one can achieve not only better resource utilization but also enjoy performance gains through data aware routing. Grid Dynamics will present a reference architecture for building a data aware routing layer on top of SGE and Oracle Coherence along with a demo that highlights the performance benefits of data aware routing. Sun HPC ClusterTools and Open MPI Update
Developing Applications for Parallelism — Ruud van der Pas
The multicore trend in microprocessor architectures turns even every laptop into a parallel system. By clustering (multicore) systems, performance is virtually unlimited, but the software needs to be adapted to take advantage of this potential.
Automatic parallelization by the compiler, as well as OpenMP are excellent ways to parallelize an application for a single parallel system.
For those that need more performance than provided by a single system, MPI is a very good programming model to develop and deploy a parallel application targeting a cluster of systems.
After a brief review of these three ways to parallelize an application, we zoom in on the features and tools provided by Sun Studio and Sun HPC ClusterTools to assist the developer writing and executing parallel applications.
Developing Applications for Performance — Ruud van der Pas
In this talk we focus on how to squeeze the maximum performance out of an application using the Sun Studio compilers and tools.
Compilers are very powerful, but also complex tools to assist the developer. A good understanding of a compiler is of great benefit to get the best performance with minimal effort.
After an overview of the architecture of the Sun Studio compilers, we give tips and tricks how to get the maximum benefit using these tools when optimizing an application.
EnginFrame Grid and Cloud Portal — Karsten Gaier
An Enterprise HPC Grid infrastructure is of immediate appeal to power users, but common engineers and researchers often have to cope with many layers of complexity and training to optimally leverage the available resources.
The Grid Portal integrated with Sun Grid Engine is the solution that allows a user-friendly, secure access to applications via a standard Web browser. By harmonizing application-specific submission, job / Grid control and end-to-end data management, the EnginFrame Grid Portal enables ubiquitous access to computing, remote visualization and collaboration. Seamless integration of cloud infrastructures transparently enhances the HPC capacity.
Execution of SGE Clusters On Top of Hybrid Clouds using OpenNebula — Constantino Vazquez
OpenNebula is an open and flexible tool that fits into existing data center environments to build any type of Cloud deployment. One of its main novelties is the support for Hybrid Cloud computing to combine local infrastructure with public cloud-based infrastructure, enabling highly scalable hosting environments. The talk will describe and demonstrate how to virtualize a SGE computing cluster on a private distributed virtual infrastructure and how to scale out the infrastructure using resources from different Cloud providers in order to meet peak or fluctuating computing demands.
Green HPC Leveraging Sun Grid Engine and scVenus — Jan Wender, Oliver Schroeder
This talk presents a project to introduce Green IT"" concepts in a commercial HPC cluster, resulting in significant savings of energy costs. Analysis of the job distribution on the Linux cluster revealed an unbalanced load pattern, coupled to the typical work cycles within the users' projects. Coupling the job scheduler (Sun Grid Engine) with power management features of the administrative software (scVenus), the cluster is now able to fully automatically power on and off systems on demand. Thus, only a subset of nodes are kept available to meet current performance requirements while idle systems will be completely shut down. The talk describes details of the various system components, the necessary modifications to the configuration, and shows the evolution of energy costs during the past year of running this concept.""
Grid Engine Best Practices — Daniel Templeton
In the Beginner's Guide to Sun Grid Engine 6.2 and Extreme Scalability Using Sun Grid Engine Software papers that have come out over the last year, a number of performance suggestions were made. This session will summarize the state of the art for cluster configuration to achieve maximum throughput and scalability.
Grid Engine Community Update — Fritz Ferstl
To kick off the Grid Engine track, the director of the Sun Grid Engine product line will talk about where the Sun Grid Engine product team, the Grid Engine open source community, and the Grid Engine technology stand and what to expect going forward.
Grid Engine Experience in Finis Terrae — Pablo Rey Mayo
CESGA is the Supercomputing Center of Galicia. We started using Grid Engine ten years ago in the x86 cluster and since then it has been installed in the different supercomputers of the center. These supercomputers had very different architectures and OS including HP-Superdome with HP-UX and HPC320 running Tru64. On April 2008, the new supercomputer Finis Terrae, a cluster of 142 nodes with 16 Itanium-2 cores and 128GB of memory per node, and a Lustre filesystem interconnected with infiniband started its production, becoming number 100 in the TOP500 with 16 Tflops of main memory, 2528 Itanium-2 cores and 19 TB of main memory. In this talk we will present our experience with the SGE software, and the developments that were introduced for the integration of different MPI flavours (HP-MPI and Intel MPI) and other utilities (qsub wrapper, jobs prioritization, management of special user requirements, ...) to facilitate the administration of the system and daily work of more than 500 users from different scientific areas. CESGA is also leading the integration of Grid Engine in the gLite middleware that is used in EGEE, the biggest grid production infrastructure in Europe and we will present an introduction of the work developed within this project.
Grid Engine Lightning Talks
This session will be a series of short talks about new and upcoming Grid Engine features:
• New Graphical Installer – Lubos Petrik, Sun
• Job Submission Verifiers – Daniel Templeton, Sun
• SGE Inspect – Jana Olivova, Sun
• Job to Core Binding BoF – Daniel Gruber, Sun
• DRMAAv2 – Peter Troeger, Humboldt University
Grid Engine Roadmap — Andy Schwierskott
This talk will present an overview of the plans for the 6.2u4 and 6.2u5 releases and well as a glimpse of what lies beyond for the Grid Engine open source project.
Grids and Clouds and Cluster HPC: Come together – Right now – Over me — Dr. Wolfgang Gentzsch
Now that we have Clouds, are Grids obsolete, are Supercomputers obsolete? Or are Clouds simply what we meant Grids to be, similar to the Grids what we meant Supercomputers to be? Does that mean that Clouds will now replace Supercomputers? You want my opinion? Then come together right now over me.
HDFS-Aware Scheduling With Grid Engine — Daniel Templeton
HDFS is a subproject of Apache Hadoop that distributes data across the nodes of the cluster. It's the foundation upon which the Hadoop Map/Reduce project is built. While Hadoop is excellent at what it does, administrators could profit from the additional benefits provided by traditional DRM technologies. The problem has been that traditional DRM technologies are unable to schedule Map/Reduce jobs according to the HDFS data.
This session will explain how to make the Grid Engine software aware of HDFS data locality, enabling the scheduling of Map/Reduce jobs according to the data. With this solution, the combination of Apache Hadoop and the Grid Engine software becomes feasible and practical.
High Performance Computing Software Solution Stacks — Art Beckman, Bill Bryce
One of the typical challenges in high performance computing is assembling and testing a reliable software stack. To make that process simpler, Sun now has two fully integrated, fully tested HPC software stack options available, with a third coming soon. This session will talk about what those stacks are, how they can benefit your organization, and where to get them.
HPC Development on OpenSolaris for Cloud Environments — Bogdan Vasiliu
The Sun HPC Software, Developer Edition for OpenSolaris gives developers a unique environment with all the tools they need to create and deploy HPC applications, such as financial, engineering, genomics, medical diagnostics, face recognition and many more compute intensive applications that cannot run in a single computer for production. It runs on OpenSolaris, combining Sun's High Performance Computing software tools (Sun Studio, HPC Cluster Tools, Sun Grid Engine) with virtualization tools like VirtualBox or VMware, and connectivity to external cloud resources.
It works with automatic, on demand, elastic environments of up to thousands of CPU cores. It meets both the increase in demand and the need for high quality of service. Based on resource demand and policies, Sun HPC Software, Developer Edition for OpenSolaris can provide elasticity from automatic connectivity to Amazon EC2.
The Sun HPC Software, Developer Edition for OpenSolaris helps developers focus on creativity. It removes the complexity inherent to HPC applications deployment. This is what this talk will show in detail.
HPC Trends: Technologies, Challenges, and Convergence — Josh Simons
High Performance Computing is about to enter a new phase that will involve fundamental changes in the way HPC applications are programmed and how HPC systems are managed and used. At the same time, the requirements of Enterprise, HPC, and Cloud IT continue to converge. These changes, which are being driven by technology trends as well as emerging gaps, will create both great opportunities and great challenges for the HPC community. Specific discussion topics will include the coming role of virtualization in HPC and the impact of new resilience requirements on vendors, deployers, and developers.
Jointly Managing Workloads and Data in the Cloud — Panel
This panel will address questions about managing workloads and their associated data in public and hybrid cloud environments. The panel will consist of members from the Grid Engine and Lustre product teams, Oracle, GigaSpaces, and Univa. It will be moderated by a representative from GridDynamics. Audience questions are encouraged.
Lustre at FZ Juelich – Status and Goals — Otto Buechner
On the 6th of August, the first Supercomputer in Juelich with a Lustre filesystem started into production. This talk will cover the configuration and first experiences with Lustre. Also plans and activities from Juelich concerning Lustre in the future will be discussed.
Lustre Clustered Metadata — Andreas Dilger
Lustre has excellent IO scalability, but currently only supports a single Metadata Server for each filesystem, limiting the metadata performance to that of a single server. Clustered Metadata will allow Lustre filesystems to scale the metadata performance across multiple server nodes. The CMD implementation will be discussed, along with the current project status.
Lustre Community — Dan Ferber
Lustre has an open source and community sponsored history. This talk reviews the various ways the Lustre engineering team and the Lustre community work together, and the Lustre information that is available on lustre.org and sun.com pages.
Lustre in Automotive — Daniel Kobras
I/O intensive simulations of crash behaviour and aerodynamic properties are core to the engineering processes in the automotive industry. In order to fully leverage the compute power of current cluster systems, parallel filesystems are a must for these types of applications. This talk summarises the experiences gained from the past two years of running Lustre installations at enterprises in the automotive sector. It not only discusses Lustre filesystems that serve as cluster scratch spaces, but also investigates the challenges and benefits of using Lustre as the main, central storage infrastructure in a commercial environment, as well as the practicality of Lustre in day-to-day administration.
Lustre Investigations at CERN — Arne Wiebalck
The particle physics laboratory CERN is currently evaluating Lustre as a potential consolidated storage solution for project space, home directories, analysis space and HSM. Rather than on performance or on scalability, the main focus of the evaluation will be on operational questions. This talk will cover CERN's storage use cases, the aspects to be looked at during the survey and the initial thoughts and findings.
Lustre Management & Monitoring Tools — Sven Trautman
The scope of this talk is to give a survey of open source tools which try to make working with Lustre easier. The talk will compare and discuss a number of tools which can be used for setting up, running and monitor Lustre file systems.
The talk will be concluded by a discussion on what is missing from current tools and what is needed to make running Lustre file systems even more enjoyable.
Lustre Performance Tips — Torben Kling-Petersen
Are your Lustre filesystem running at peak efficiency? Are you using you hardware in the most optimal way? Have you tuned the settings to get the most out of your storage solution?
This presentation is a discussion on what the Sun Lustre benchmarking specialists do to get the most out of a solution. Tips on filesystem layout, settings and tweaks. Do's and Don'ts as well as which tuning parameters that have the greatest effect.
Lustre Quota: Current and Future — Johann Lombardi
This talk reviews the current support in Lustre for quotas, and what is planned for the future.
Lustre Roadmap — Dan Ferber
This talk welcomes people to the Lustre track. It includes a brief overview of what Lustre does as well as the current version and plans for upcoming Lustre releases and plans.
Lustre, SGE, and Physics Codes — Rajmund Krivec
The Department of Theoretical Physics of J. Stefan Institute, Ljubljana, is running a Lustre over Infiniband installation centered on a pair of 16-CPU X4600s, two MDS, and a 6 TB X4500 OSS. Lustre FS is exported to other machines via NFS, and managed through the Sun Grid Engine. While the system I/O throughput is as expected, in openMP mode, and we have some elegant, almost single-directive, real physics code examples of openMP parallelization, we have run into some stability issues (timeouts, client eviction) which have not yet been fully resolved.
Lustre, ZFS, and Data Integrity — Andreas Dilger
An overview of the upcoming ZFS functionality will be provided, including improved scalability, hybrid storage pools, snapshots, and data checksumming. Designs for efficiently integrating the Lustre network data integrity with the ZFS data checksumming will also be discussed.
Meet the Grid Engine Team — The Grid Engine Team
Have you ever wondered who crei is? Or ernst? Or lubos? This session is your opportunity to meet the engineers behind the product. After the engineers introduce themselves and what they do in the product team, there will be the opportunity for questions and discussion.
Meet the Lustre Team — The Lustre Team
Just before we break for the day and get ready for the Gala Dinner, come and meet the Lustre engineers who are attending the conference this week, say hello, and ask questions.
Open Cloud Framework-Open Standards for the Cloud Community — Thijs Metsch
Most of today's cloud offerings support infrastructure as a service (IaaS) based models. Elasticity rules and intelligent SLA management are add on features which are essential for large deployments. Such deployments have a demand for portability, integration and interoperation of cloud services. This talk will present one of the major EU projects: RESERVOIR in which an open cloud framework is under development. It supports the deployment of virtual workloads (which includes virtual machines as well as Java services) over standardized interfaces. The OGF OCCI emerging standard with it's features will be presented in conjunction with the RESERVOIR project.
Oracle Coherence: The Data Grid in your Compute Grid — Brian Oliver
This talk will provide a technical introduction to Oracle Coherence Data Grid solution, it's origin and evolution, how it works, how it allows data and compute intensive applications to reliably scale horizontally and why it's increasingly becoming the standard data back bone of compute grids for commercial enterprises around the world. Included in the talk will be a live demonstration of fault-tolerant horizontal scaling, together with an examination of using Data Grids such as Coherence as a Compute Grid.
ProjectFortress: Run Your Whiteboard — Josh Simons, Sun
The ProjectFortress programming language is targeted at domain experts. We want them to be able to express their algorithms in whatever terms they would use when writing on a whiteboard discussing their problems with a colleague. This talk will give you an overview of ProjectFortress and some examples to give you a feel for programming in the language. We have an open source implementation which you are free to download and experiment with.
Real-Time Analytics on the Cloud — Nati Shalom
Real time analysis application requires a large pool of compute resources. Building such an application for risk analysis, profit & loss analysis requires pre-investment in building large data centers to support such a scale. Setting up such an environment is therefore a luxury that only a few big enterprise firms can afford.
Cloud computing provides a pool of machines that can be consumed on demand at a relatively low cost. This opens up new opportunities for running such applications by smaller firms. It also enables those companies to offer their solution as a service. This concept can revolutionize the way analytic application is built. In this session we will learn the lessons of running a risk analysis application as a service on Amazon cloud using the GigaSpaces Platform as a Service (PaaS). We will review a real life case study (Primatics) and examine how they dealt with the challenges of building such an application and handling security, scaling and multi-tenancy concerns etc.
Sample Lustre Performance Data — Dan Ferber
Beginning this Autumn Dan’s team will be responsible for collecting Lustre performance data for external use as well as internal analysis and improvements. This talk reviews some current Lustre performance data and plans for additional data collection and analysis.
SGE in EDA: Helping Engineers Design In The Cloud — Bill Bryce
Electronics design automation (EDA) companies can substantially increase the flexibility of their large throughput computing environment with private cloud computing. Univa's HPC Cloud solution - with Sun Grid Engine, Oracle Enterprise Linux, Oracle VM and Univa's UniCloud 2.0 - closes the loop between job placement, provisioning and policy management. Thus, customers can optimize priority work, fully utilize existing clusters and provision capacity on demand in a public cloud. In such an environment, jobs are always run on the most eligible resource available and can be migrated, checkpointed or sent to a data center or external cloud where power costs less or spare licenses are available. This solution redefines 'Moore's Law' into 'Mores Law' - as in, Get More For Less.
SGE: Why Clouds? – A Business Model View — Miha Ahronovitz
Sun Grid Engine can play a key role as a building block for hybrid clouds. The presentation describes high level the business model of the classical grid and its challenges to offer constant level of service. Then, we describe how resources are utilized in an elastic model that combines local cloud resources with external node instances from public clouds, to create a Hybrid Cloud. The "no need to over design", substantial power savings, the ability to pass internal Cloud billings to end customers, make the hybrid cloud a profit center inside an organization.
Sun Development Tools for High-Performance Computing — Ruud van der Pas
Sun Studio combined with Sun HPC ClusterTools provides a comprehensive tool set to develop serial and parallel applications. In this introductory talk we focus on those components and features of interest to the development of HPC applications.
After an overview of the product features, we zoom in on several components, demonstrating how these tools can be used to assert the performance, identify bottlenecks and find out what optimizations the compiler has performed. We also present and discuss the support for the development of parallel applications using Automatic Parallelization by the compiler, as well as explicit parallelization using OpenMP and/or MPI.
Sun HPC ClusterTools and Open MPI Update — Terry Dontje
Terry Dontje of the ClusterTools engineering team will be reviewing the features released in CT 8.2 plus new features to be included in upcoming releases for CT and Open MPI.
Sun HPC System Solutions — Constantin Gonzalez
Every software touches hardware, eventually. Therefore, we'll have a look at the current state of Sun's HPC hardware products, from individual systems to blade servers to the Sun Constellation System that delivers the biggest density in x86 compute power, networked through a powerful QDR Infiniband switch. We'll also discuss some storage hardware options that go well with Lustre as well as the Sun Unified Storage Systems that provide easy-to-use NAS capabilities for smaller installations.
Sun Lustre Storage Server — Torben Kling-Petersen
Sun Microsystems reference architecture for a Lustre based storage solution is built on standard servers and JBODs configured for optimal perfromance and reliability while keeping the total cost of ownership down. This presentation will cover the components of the current version as well as providing a preview of the forthcomming update.
Sun Solutions for Optimizing MD Nastran Performance — Dale Layfield
MSC Software and Sun Microsystems have worked together for two decades to optimize MSC's major applications, including MD Nastran, on Sun Systems. This talk will outline the recent collaboration between MSC.Software and Sun in the following two areas:
1.) Sun Grid Engine and MD Nastran integration: The presentation will included a detailed discussion of the recently completed project on the tight-integration of Sun Grid Engine (SGE) and MD Nastran. The presentation will describe the recent series of how-to"" blogs on this MD Nastran and SGE integration effort (posted on Sun's blogs.sun site). The ""how-to"" blogs lay out in a step-by-step format how to configure Sun Grid Engine to work effectively with MD Nastran. Making use of MD Nastran's existing job queueing interface capabilities as well as MSC's resource estimation tool (the ""ESTIMATE"" program) this presentation will give an overview of how to configure typical SGE job queues for MD Nastran, submit serial and parallel jobs, and how to manage resource utilization (memory, disk space, and license tokens) to get optimum throughput on a compute cluster. And finally, a look at how MSC's SimManager, an Enterprise Simulation Management system, works together with MD Nastran and Sun Grid Engine to manage simulation processes, resources, and information.
2.) A new Sun HPC BluePrint Article on MD Nastran--""A comprehensive tuning and hardware configuration ""best practice"" document for optimizing MD Nastran performance"": This part of the presentation will include a detailed discussion of recent MD Nastran performance studies done on a variety of Sun hardware configurations (including results on the latest Nehalem processors, memory configurations, Intel Solid State Devices (SSDs), Sun Storage 7xxx Unified Storage systems, and the Lustre filesystem). The talk will also include detailed hardware and software configuration recommendations that can help reduce MD Nastran elapsed times by 50% or more. This part of the talk will also describe how to implement three key recommendations offered in the document: (1) How to use a hierarchical storage access method for optimum I/O performance (2) How to run more jobs (than ever before) on your compute cluster (3) How SGE can be configured to fully realize the potential of the first two recommendations.
The Time is Now for HPC Cloud Computing: Technology Overview and Case Study — Gary Tyreman
Cloud may be here"" – but is it really here in production, or are most implementations merely low-level proofs of concept for less-than-realistic use cases? Find out in this presentation where we'll detail a case study in which one market leader substantially increased the flexibility of their large throughput computing environment with a Univa Private Cloud. In such an environment, jobs are always run on the most eligible resource available and can be live-migrated, checkpointed or sent to a data center or external cloud with zero downtime and no end user impact. Think VMware VMotion with policy-enabled dynamic resource management for technical computing in a hybrid cloud scenario, where your most high-performance (and high-value) apps are optimally utilized across a pool of virtual and physical shared resources.""
Trends in High-Performance Computing — Chris Dagdigian
BioTeam Inc. principal Chris Dagdigian works as an independent consultant specializing in enabling high performance research computing primarily in pharmaceutical, biotech and other life science environments. This talk will cover recent projects and observed trends and will be a preview of his address at the 2009 Bio-IT World Conference & Expo in Hannover, Germany on October 6-8.
Unicluster + Sun Grid Engine — Bill Bryce
Univa is the leading provider of HPC systems management software and an expert in LSF migration. Our HPC products embed Sun Grid Engine (SGE) as the core scheduling component and we offer a single path for support of SGE environments. Univa is also an award-winning innovator in cloud computing enablement, building private clouds from Grid Engine clusters and extending HPC environments to public clouds. Come learn more about what Univa does and how we benefit Sun customers and partners.
Using Sun Studio in HPC & Education — Bernd Dammann
In this presentation we show how Sun Studio is used at the Technical
University of Denmark, both as a tool to teach basic principles in
High-Performance Computing, as well as a powerful set of tools in the
development and deployment of HPC applications.
We will show a number of different examples, to illustrate some of the
aspects of Sun Studio, such as compiler options and features, as well
as the analysis tools for serial and parallel programming.
What's New in Grid Engine 6.2 — Daniel Templeton
A lot has happened since the 6.1 release in 2007. With the 6.2, 6.2u2_1, and 6.2u3 release, we've introduced a number of new and interesting features. This talk will take a survey of what's new with the 6.2 releases and why these new features are important.
What's Next For Grid Engine? — Panel
In this panel session, you will have the opportunity to tell the Grid Engine product team about what direction you'd like to see the product taking and what features are important to you. On the panel will be the director, the engineering manager, the product manager, and the community manager.
Sign up or Log in to add a comment or watch this page.