Profiling/Timeseries

Here's what we're doing:

• Running CPU intensive code
• cpu, memory, threads, file size
• where in structure, depth, iterations, etc.

Architecture:

• Use Zmq
• Zmq client is code running on device, populating message queue
• Zmq server is server listening for messages, filtering takes place on server

N queens problem:

• searching for number of ways to arrange n queens on the chessboard
• code instrumentation: current portion of tree (vector of queen placements), number of solutions found
• profiling: netdata on host machine