Posts Implementing DAG in Python

Implementing DAG in Python

Workflow Oriented

Why are DAGs good for task dependencies? The ‘G’ in DAG is ‘Graph’. A Graph is a collection of nodes and edges that connect the nodes. For our purposes, each node would be a task, and each edge would be a dependency. The ‘D’ in DAG stands for ‘Directed’. This means that each edge has a direction associated with it. So we can interpret the edge (a,b) as meaning that b depends on a, whereas the edge (b,a) would mean a depends on b. The ‘A’ is ‘Acyclic’, meaning that there must not be any closed loops in the graph. This is important for dependencies, because if a loop were closed, then a task could ultimately depend on itself, and never be able to run. If your workflow can be described as a DAG, then it is impossible for your dependencies to cause a deadlock.


This post is licensed under CC BY 4.0 by the author.