[CX] Improved statistics collection

For the Analytics Service the web console currently displays information about

• the heap used,
• the system load,
• the thread count,
• the garbage collection count and time, and
• the bytes read and written (Linux only)

in the "Server Resources" section and about

• the number of sync'd records/sec,
• the number of failed records (e.g. binary values), and
• the total number of sync'd records

in the "Analytics Stats" section.
We need to review this list and identify additional metrics that can be used to understand the state and behavior of the system. Possible additions could be the DCP queue size or more detailed information about the resource utilization on Analytics (or other) nodes.