Debian Science Distributed Computing packages

The Modules package provides for the dynamic modification of a user's environment via modulefiles. Each modulefile contains the information needed to configure the shell for an application. Once the Modules package is initialized, the environment can be modified dynamically on a per-module basis using the module command which interprets modulefiles. Typically modulefiles instruct the module command to alter or set shell environment variables such as PATH, MANPATH, etc. modulefiles may be shared by many users on a system and users may have their own collection to supplement or replace the shared modulefiles. The modules environment is common on SGI/Crays and many workstation farms.

Grid Engine é um software que facilita a "distribuição de recursos gerenciamento" (Distributed Resource Management - DRM). Muito mais do que simples ferramentas de balanceamento de carga ou mecanismos de agendamento em lote, o software DRM normalmente fornece as seguintes funcionalidades essenciais em grandes conjuntos de recursos distribuídos:

• Alocação baseada em política de recursos distribuídos (tempo de CPU, licenças de software, etc.)
• Enfileiramento e agendamento em lote
• Suporta diversos hardwares de servidor, sistemas operacionais e arquiteturas
• Balanceamento de carga e execução remota de tarefas
• Estatísticas detalhadas da contabilidade de jobs
• Recursos especificados por usuários(as) de maneira detalhada
• Suspender/retomar/migrar jobs
• Ferramentas para relatar o status do Job/Host/Cluster
• Matrizes de jobs
• Integração e controle de jobs paralelos

Este pacote contém os utilitários de usuário(a) do Grid Engine.

Grid Engine is software that facilitates "distributed resource management" (DRM). Far more than just simple load-balancing tools or batch scheduling mechanisms, DRM software typically provides the following key features across large sets of distributed resources:

• Policy based allocation of distributed resources (CPU time, software licenses, etc.)
• Batch queuing & scheduling
• Support diverse server hardware, OS and architectures
• Load balancing & remote job execution
• Detailed job accounting statistics
• Fine-grained user specifiable resources
• Suspend/resume/migrate jobs
• Tools for reporting Job/Host/Cluster status
• Job Arrays
• Integration & control of parallel jobs

This package contains the Grid Engine execution server.

Grid Engine is software that facilitates "distributed resource management" (DRM). Far more than just simple load-balancing tools or batch scheduling mechanisms, DRM software typically provides the following key features across large sets of distributed resources:

• Policy based allocation of distributed resources (CPU time, software licenses, etc.)
• Batch queuing & scheduling
• Support diverse server hardware, OS and architectures
• Load balancing & remote job execution
• Detailed job accounting statistics
• Fine-grained user specifiable resources
• Suspend/resume/migrate jobs
• Tools for reporting Job/Host/Cluster status
• Job Arrays
• Integration & control of parallel jobs

This package contains the Grid Engine master server.

Grid Engine é um software que facilita o "gerenciamento de recursos distribuídos" (Distributed Resource Management - DRM). Mais do que simples ferramentas de balanceamento de carga ou mecanismos de agendamento em lote, o software DRM geralmente provê os seguintes recursos-chaves em conjuntos extensos de recursos distribuídos:

• Alocação baseada em política de recursos distribuídos (tempo de CPU, licenças de software, etc.)
• Enfileiramento e agendamento em lote
• Suporte a diversos hardwares de servidor, SOs e arquiteturas
• Balanceamento de carga e execução remota de trabalhos
• Estatísticas detalhadas da contabilidade de trabalhos
• Recursos específicos de usuário(a) de maneira detalhada
• Suspensão/resumo/migração de trabalhos
• Ferramentas para relatar o status de Job/Host/Cluster
• Matrizes de Trabalhos
• Integração e controle de trabalhos paralelos

Este pacote contém a interface gráfica de administração do Grid Engine.

This package aids upgrades of existing htcondor installations to the package name "condor". The package is empty and it can safely be removed.

O IPython pode ser usado como substituto para o interpretador de comandos interativos padrão do Python, ou pode ser usado como um ambiente de trabalho completo para computação científica (como Matlab ou Mathematica) quando usado em conjunto com as ferramentas científicas e numéricas padrão do Python. Suporta introspecção de objetos dinâmicos, numeração de entrada/saída de comandos, sistema de macros, registros de atividades de sessão, restauração de sessão, acesso completo ao interpretador de comandos do sistema, relatórios detalhados e coloridos de rastreio, inserção automática de parênteses, inclusão automática de aspas e pode ser embutido em outros programas Python.

Este pacote contém o terminal final do interpretador de comandos para Python 3.

NorduGrid is a collaboration aiming at development, maintenance and support of the middleware, known as the Advanced Resource Connector (ARC).

This client package contains all the CLI tools that are needed to operate with x509 proxies, submit and manage jobs and handle data transfers.

Open MPI é um projeto que combina tecnologias e recursos de vários outros projetos (FT-MPI, LA-MPI, LAM/MPI e PACX-MPI) para construir a melhor biblioteca MPI disponível. Uma implementação compatível com MPI-3.1 completamente nova, a Open MPI oferece vantagens para fornecedores(as) de sistemas e software, desenvolvedores(as) de aplicativos e pesquisadores(as) de ciência da computação.

Recursos:

• Conformidade total com os padrões MPI-3.1
• Segurança de thread e simultaneidade
• Processo dinâmico de desova
• Alto desempenho em todas as plataformas
• Gerenciamento de trabalho confiável e rápido
• Tolerância a falhas de rede e processo
• Suporta heterogeneidade de rede
• Uma única biblioteca suporta todas as redes
• Instrumentação em tempo de execução
• Muitos programadores de trabalho são suportados
• Mensagens de erro internacionalizadas
• Design baseado em componentes, APIs documentadas

Este pacote contém os programas utilitários Open MPI.

SCOOP (Scalable COncurrent Operations in Python) is a distributed task module allowing concurrent parallel programming on various environments, from heterogeneous grids to supercomputers.

This package contains the Python3 modules.

This is a collection of software that help users to share resources in a complex, heterogeneous, and unreliable computing environment. This includes:

• Chirp: A personal filesystem and I/O protocol that allows unprivileged users to share space securely, efficiently, and conveniently. When combined with Parrot, Chirp allows users to create custom wide-area distributed filesystems.
• Parrot: A transparent user-level virtual filesystem that allows any ordinary program to be attached to a remote storage device such as an FTP server or a Chirp server.
• Makeflow: A workflow system for parallel and distributed computing that uses a language very similar to Make.
• Work Queue: A system and API for building master-worker style programs that scale up to thousands of processors.
• All Pairs: A computational abstraction for running very large Cartesian products.
• Wavefront: A computational abstraction for running very large dynamic programming problems.
• The Fault Tolerant Shell: A high-level programming language that allows users to combine the ease of shell scripting, the power of distributed programming, and the precision of compiled languages. Basically, parallel programming and exception handling for scripts.

The High Performance Computing (HPC) Challenge benchmark runs a suite of 7 tests that measure the performance of CPU, memory and network for HPC clusters. Amongst others, it includes the High-Performance LINPACK (HPL) benchmark, used by the Top500 ranking (http://www.top500.org/).

MPI for Python (mpi4py) provides bindings of the Message Passing Interface (MPI) standard for the Python programming language, allowing any Python program to exploit multiple processors.

mpi4py is constructed on top of the MPI-1/MPI-2 specification and provides an object oriented interface which closely follows MPI-2 C++ bindings. It supports point-to-point (sends, receives) and collective (broadcasts, scatters, gathers) communications of any picklable Python object as well as optimized communications of Python object exposing the single-segment buffer interface (NumPy arrays, builtin bytes/string/array objects).

PSOM is a lightweight library to manage complex multi-stage data processing. A pipeline is a collection of jobs, i.e. Matlab or Octave code with a well identified set of options that are using files for inputs and outputs. To use PSOM, the only requirement is to generate a description of a pipeline in the form of a simple Matlab/Octave structure. PSOM then automatically offers the following services:

• Run jobs in parallel using multiple CPUs or within a distributed computing environment.
• Generate log files and keep track of the pipeline execution. These logs are detailed enough to fully reproduce the analysis.
• Handle job failures: successful completion of jobs is checked and failed jobs can be restarted.
• Handle updates of the pipeline: change options or add jobs and let PSOM figure out what to reprocess.