Feature: New Metrics endpoint for CRN & CNN

[1yam]

Problem:

The current approach for getting those information is to fetch the whole list of nodes using ranges of post messages, which means pulling a whole lot of needless information (35mb for a week of CRN metrics).

Solutions:

Implementing two API endpoints to query CRN & CNN metrics in the Pyaleph database.

1. /api/v0/core/{node_id}/metrics

{
    "metrics": {
        "item_hash": [<string>, <string>, ...],
        "measured_at": [<UNIX timestamp>, <UNIX timestamp>, ...],
        "base_latency": [<float>, <float>, ...],
        "base_latency_ipv4": [<float>, <float>, ...],
        "metrics_latency": [<float>, <float>, ...],
        "aggregate_latency": [<float>, <float>, ...],
        "file_download_latency": [<float>, <float>, ...],
        "pending_messages": [<int>, <int>, ...],
        "eth_height_remaining": [<int>, <int>, ...],
    }
}

3. /api/v0/compute/{node_id}/metrics

{
    metrics: {
        item_hash: [<string>, <string>, ...],
        measured_at: [<UNIX timestamp>, <UNIX timestamp>, ...],
        base_latency: [<float>, <float>, ... ],
        base_latency_ipv4: [<float>, <float>, ...],
        full_check_latency: [<float>, <float>, ...],
        diagnostic_vm_latency: [<float>, <float>, ...],
    }
}

Params:

• start_date: float
• end_date: float
• sort: ASC or DESC

[MHHukiewitz]

One typo, otherwise LGTM