Investigate outcome of stress test

[muXxer]

Checkout logs/captures and find the bottlenecks causing the highest resource usage at the node's breakpoint.

[muXxer]

First investigations:

hornet:

• keeping the latest cone completely in RAM would reduce parents and childrens walk workload by a lot
• signature checks are quite heavy and currently single threaded

inx-coordinator:

• almost no CPU load even at 1000 BPS

hornet node of inx-coordinator:

• high CPU load because of garbage collection and whiteflag
• its double the CPU load of other nodes (whiteflag is done twice per milestone)

inx-dashboard:

• creates unnecessary CPU load if no client is connected to fill the visualizer