Node statistics and performance values¶

The diameter stack collects statistical data on its performance.

Statistics are provided on three different levels:

Peer statistics¶

Statistics are collected individually for each configured peer, accessible through the statistics instance attribute, which holds an instance of PeerStats.

Peer statistics record the following values:

avg_response_time

Average time in seconds spent processing a request, recorded individually for each received request type. Contains a dictionary, with request types as keys and average response times as values, e.g.:

{"Capabilities-Exchange": 1.0, "Device-Watchdog": 0.000157}

The value is calculated from the time spent processing the last 1024 requests.

avg_response_time_overall

Average time in seconds spent processing any request received through the peer.

The value is calculated from the time spent processing the last 1024 requests.

processed_req_per_second

The rate of requests processed, per second, recorded individually for each received request type. Contains a dictionary, with request types as keys and rate of processed requests as values. E.g.:

{"Capabilities-Exchange": 1.0, "Device-Watchdog": 3.0}

The rate is calculated by adding the total sum of time spent processing the last 1024 requests, separately for each request type, and dividing the sums by the amount of requests.

The request rate is likely to be very high, if the peer receives low traffic.

processed_req_per_second_overall

The overall rate of requests processed, per second, for the entire peer.

Similar to processed_req_per_second, the rate is calculated by adding the total sum of time spent on processing the last 1024 requests, for any request type, and diving the sum by the amount of requests.

Node statistics¶

The diameter node collects statistics for all of its peers and calculates cumulated values. The node statistics are available as the statistics instance attribute, which returns an instance of NodeStats.

The node statistics record following values:

avg_response_time

Average response time, recorded individually for each request type. Identical to the same attribute in peer statistics, except calculated cumulated over all peers.

avg_response_time_overall

Average response time for the entire node. Identical to the same attribute in peer statistics, except calculated cumulated over all peers.

processed_req_per_second

Rate of requests processed per second. Identical to the same attribute in peer statistics, except calculated cumulated over all peers.

processed_req_per_second_overall

Rate of requests processed per second. Identical to the same attribute in peer statistics, except calculated cumulated over all peers.

received_req_counters

A list containing three integer values, indicating the exact amount of request received by the node in the last minute, five minutes and the last 15 minutes.

sent_result_code_range_counters

A dictionary containing exact values of sent result code types. The dictionary holds result code ranges as keys, and lists of integers as values. Each result code range has the format of "1xxx", "2xxx" etc. Each value is a list with three integers; the amount of sent result codes for the last minute, last 5 minutes and last 15 minutes. E.g.:

{"2xxx": [150,321,321], "4xxx": [58,103,103], "5xxx": [1,1,1]}

Historical statistics¶

The node, as long as it is running, takes a snapshot of the Node.statistics return value every 60 seconds and stores it as a serialised dictionary in the statistics_history instance attribute. The history is a list (a deque) of dictionaries and holds 1440 entries, i.e. historical values for the past 24 hours.

Values of each dictionary correspond to attribute names of NodeStats, with one timestamp key added, containing the UNIX timestamp of the snapshot.

[
    {
        "avg_response_time": {
            "Capabilities-Exchange": 1.0,
            "Credit-Control": 0.033816425120772944,
            "Device-Watchdog": 0.3333333333333333
        },
        "avg_response_time_overall": 0.029674902627634447,
        "processed_req_per_second": {
            "Capabilities-Exchange": 1.0,
            "Credit-Control": 29.571428571428573,
            "Device-Watchdog": 3.0
        },
        "processed_req_per_second_overall": 30.714285714285715,
        "received_req_counters": [209, 213, 213],
        "sent_result_code_range_counters": {
            "2xxx": [166, 171, 171],
            "4xxx": [44, 44, 44]
        },
        "timestamp": 1705507856
    },
    {
        "avg_response_time": {
            "Capabilities-Exchange": 1.0,
            "Credit-Control": 0.03357314148681055,
            "Device-Watchdog": 0.3333333333333333
        },
        "avg_response_time_overall": 0.028846585890826056,
        "processed_req_per_second": {
            "Capabilities-Exchange": 1.0,
            "Credit-Control": 29.785714285714285,
            "Device-Watchdog": 3.0
        },
        "processed_req_per_second_overall": 30.357142857142858,
        "received_req_counters": [209, 423, 423],
        "sent_result_code_range_counters": {
            "2xxx": [150, 321, 321],
            "4xxx": [58, 103, 103],
            "5xxx": [1, 1, 1]
        },
        "timestamp": 1705507916
    }
]