pgiri/pycos

Concurrent, Asynchronous, Distributed, Communicating Tasks with Python

pycos

Note

Full documentation for pycos is now available at pycos.org.

pycos is a Python framework for asynchronous, concurrent, network / distributed programming, distributed computing with very light weight tasks and message passing.

Unlike with other asynchronous frameworks, programs developed with pycos have same logic and structure as programs with threads, except for a few syntactic changes - mostly using yield with asynchronous completions that give control to pycos's scheduler, which interleaves executions of tasks, similar to the way an operating system executes multiple processes. In addition, pycos has many additional features, including message passing for communication, distributed computing/programming etc.

Unlike threads, creating tasks with pycos is very efficient. Moreover, with pycos context switch occurs only when tasks use yield (typically with an asychronous call), so there is no need for locking and there is no overhead of unnecessary context switches.

pycos works with Python versions 2.7+ and 3.1+ on Linux, Mac OS X and Windows; it may work on other platforms (e.g., FreeBSD and other BSD variants) too.

Features

• No callbacks or event loops! No need to lock critical sections either,
• Efficient polling mechanisms epoll, kqueue, /dev/poll, Windows I/O Completion Ports (IOCP) for high performance and scalability,
• Asynchronous (non-blocking) sockets and pipes, for concurrent processing of I/O,
• SSL for security,
• Asynchronous locking primitives similar to Python threading module,
• Asynchronous timers and timeouts,
• Message passing for (local and remote) tasks to exchange messages one-to-one with Message Queue Pattern or through broadcasting channels with Publish-Subscribe Pattern,
• Location transparency with naming and locating (local and remote) resources,
• Remote Pico Service (RPS) for defining services that remote clients can run as tasks (with possibly message passing to communicate).
• Distributing computation components (code and data) for execution of distributed communicating processes, for wide range of use cases, covering SIMD, MISD, MIMD system architectures at the process level, web interface to monitor cluster/application status/performance; in-memory processing, data streaming, real-time (live) analytics and cloud computing are supported as well,
• Monitoring and restarting of (local or remote) tasks, for fault detection and fault-tolerance,
• Hot-swapping of task functions, for dynamic system reconfiguration.
• Thread pools with asynchronous task completions, for executing (external) synchronous tasks, such as reading standard input.

Dependencies

pycos is implemented with standard modules in Python.

If psutil is available on nodes, node availability status (CPU, memory and disk) is sent in status messages, and shown in web browser so node/application performance can be monitored.

Under Windows efficient polling notifier I/O Completion Ports (IOCP) is supported only if pywin32 is available; otherwise, inefficient select notifier is used.

Installation

To install pycos, run:

python -m pip install pycos

Authors

• Giridhar Pemmasani

Links

• Documentation is at pycos.org.
• GitHub (Code Repository).