3

I would like to refer to the question that can be found at this address :

Running alloy analyzers in parallel

Is there any ongoing research or conclusion reached on the decomposition of Alloy models, in order to allow a more optimal analysis of models ?

This interests me greatly.

Community
  • 1
  • 1
fifou
  • 65
  • 6

1 Answers1

2

I am very interesting in this topic too. Maybe we can think about it like this. When starting Alloy engine to solve a constraint, we can call a function from Alloy and ask it to solve one constraint. I think we can call this function in cluster mode, and ask each node to solve one constraint. Then, we can compute the subset of the results from each node. See here for example: http://alloy.mit.edu/alloy/code/ExampleUsingTheCompiler.java.html

I am not sure we can work like this, but it's worthy to think about it in MapReduce like framework.

Chong Tang
  • 2,066
  • 15
  • 12