GetNetworkStats improvment: timeout & caching

[bchamagne]

Description

The GetNetworkStats message is called by many nodes at the same time, the computation it does is heavy. I added a JobCache so the computation is only done once (cache reset after 30s).
I also added a 5s timer which is bigger than the 3s default.
Fixes #1325

Type of change

• Bug fix (non-breaking change which fixes an issue)

How Has This Been Tested?

This can only be tested in stress condition. I'll setup some test environment tomorrow.

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• Any dependent changes have been merged and published in downstream modules

[Neylix]

I think we can have a better way to handle this heavy calculation process.
The StatsCollector module is already doing one part of the network stats calculation by requesting only once the stats for all subsets.
We could reuse this module to also calculate only once the network patch for all the GetNetworkStats message a node receive.

Also those calculation/request are done only by the beacon summary node at summary time (00:00 currenlty) and are usefull before the self repair time (00:05 currently)

So we could use the StatsCollector module as an orchestrator that will start the 2 task (request/calculation) at summary time (using PubSub event) and delegate the calculation to JobCache module. Then store all client that request an information and return the result from the JobCache. Then when the self repair time is triggered (using Pubsub event) the module can clean the JobCache as the result will not be relevant anymore.