An auxiliary library for programming distributed algorithms?

Question

When you code a distributed algorithm, do you use any library to model abstract things like processor, register, message, link, etc.? Is there a library that does this?

I think, for example, of self-tuning algorithms, such as self-stabilizing minimum spanning tree algorithms.

Answer 1

Here is a DVM system that can be used to implement various distributed algorithms. It runs on top of MPI .

However, this is more for matrix-oriented scientific algorithms, where the distribution is performed in terms of data blocks. I had a short experience of use - it is much more convenient than direct use of MPI, and allows you to use much more readable and supported code.

Answer 2

You can check OpenMP .

Answer 3

, Apache:

• Apache Hadoop
• Apache Zookeeper - , . .

There is also the Java Parallel Processing Framework , which abstracts all machines and their composition, allowing you to focus on processing.

Depending on your needs.