Take into account system and network load during query planning

[rubensworks]

Issue type:

• ➕ Feature request

---

Description:

Our new join bus exposes metadata on CPU/memory usage and network delays.
This information may be used to create intelligent mediators to select join actors based on system and network load.

For example, a hash-join may only be used if memory usage is low, and CPU usage is high, while otherwise a nested-loop-join would be used.
Another example is the possibility of tracking the average HTTP delay of a TPF server, and modifying the join algorithm based on that (=join mediation for adaptive query processing over ldfs).

Do papers on resource constrained join mediation already exist?

[github-actions]

Thanks for the suggestion!