1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
//! Observes metrics in JSON format.
//!
//! Metric scopes are used to provide the hierarchy of metrics.  As an example, for a
//! snapshot with two metrics — `server.msgs_received` and `server.msgs_sent` — we would
//! expect to see this output:
//!
//! ```c
//! {"server":{"msgs_received":42,"msgs_sent":13}}
//! ```
//!
//! If we added another metric — `configuration_reloads` — we would expect to see:
//!
//! ```c
//! {"configuration_reloads":2,"server":{"msgs_received":42,"msgs_sent":13}}
//! ```
//!
//! Metrics are sorted alphabetically.
//!
//! ## Histograms
//!
//! Histograms are rendered with a configurable set of quantiles that are provided when creating an
//! instance of `JsonBuilder`.  They are formatted using human-readable labels when displayed to
//! the user.  For example, 0.0 is rendered as "min", 1.0 as "max", and anything in between using
//! the common "pXXX" format i.e. a quantile of 0.5 or percentile of 50 would be p50, a quantile of
//! 0.999 or percentile of 99.9 would be p999, and so on.
//!
//! All histograms have the sample count of the histogram provided in the output.
//!
//! ```c
//! {"connect_time_count":15,"connect_time_min":1334,"connect_time_p50":1934,
//! "connect_time_p99":5330,"connect_time_max":139389}
//! ```
//!
#![deny(missing_docs)]
use hdrhistogram::Histogram;
use metrics_core::{Builder, Drain, Key, Label, Observer};
use metrics_util::{parse_quantiles, MetricsTree, Quantile};
use std::collections::HashMap;

/// Builder for [`JsonObserver`].
pub struct JsonBuilder {
    quantiles: Vec<Quantile>,
    pretty: bool,
}

impl JsonBuilder {
    /// Creates a new [`JsonBuilder`] with default values.
    pub fn new() -> Self {
        let quantiles = parse_quantiles(&[0.0, 0.5, 0.9, 0.95, 0.99, 0.999, 1.0]);

        Self {
            quantiles,
            pretty: false,
        }
    }

    /// Sets the quantiles to use when rendering histograms.
    ///
    /// Quantiles represent a scale of 0 to 1, where percentiles represent a scale of 1 to 100, so
    /// a quantile of 0.99 is the 99th percentile, and a quantile of 0.99 is the 99.9th percentile.
    ///
    /// By default, the quantiles will be set to: 0.0, 0.5, 0.9, 0.95, 0.99, 0.999, and 1.0.
    pub fn set_quantiles(mut self, quantiles: &[f64]) -> Self {
        self.quantiles = parse_quantiles(quantiles);
        self
    }

    /// Sets whether or not to render the JSON as "pretty."
    ///
    /// Pretty JSON refers to the formatting and identation, where different fields are on
    /// different lines, and depending on their depth from the root object, are indented.
    ///
    /// By default, pretty mode is not enabled.
    pub fn set_pretty_json(mut self, pretty: bool) -> Self {
        self.pretty = pretty;
        self
    }
}

impl Builder for JsonBuilder {
    type Output = JsonObserver;

    fn build(&self) -> Self::Output {
        JsonObserver {
            quantiles: self.quantiles.clone(),
            pretty: self.pretty,
            tree: MetricsTree::default(),
            histos: HashMap::new(),
        }
    }
}

impl Default for JsonBuilder {
    fn default() -> Self {
        Self::new()
    }
}

/// Observes metrics in JSON format.
pub struct JsonObserver {
    pub(crate) quantiles: Vec<Quantile>,
    pub(crate) pretty: bool,
    pub(crate) tree: MetricsTree,
    pub(crate) histos: HashMap<Key, Histogram<u64>>,
}

impl Observer for JsonObserver {
    fn observe_counter(&mut self, key: Key, value: u64) {
        let (levels, name) = key_to_parts(key);
        self.tree.insert_value(levels, name, value);
    }

    fn observe_gauge(&mut self, key: Key, value: i64) {
        let (levels, name) = key_to_parts(key);
        self.tree.insert_value(levels, name, value);
    }

    fn observe_histogram(&mut self, key: Key, values: &[u64]) {
        let entry = self
            .histos
            .entry(key)
            .or_insert_with(|| Histogram::<u64>::new(3).expect("failed to create histogram"));

        for value in values {
            entry
                .record(*value)
                .expect("failed to observe histogram value");
        }
    }
}

impl Drain<String> for JsonObserver {
    fn drain(&mut self) -> String {
        for (key, h) in self.histos.drain() {
            let (levels, name) = key_to_parts(key);
            let values = hist_to_values(name, h.clone(), &self.quantiles);
            self.tree.insert_values(levels, values);
        }

        let result = if self.pretty {
            serde_json::to_string_pretty(&self.tree)
        } else {
            serde_json::to_string(&self.tree)
        };
        let rendered = result.expect("failed to render json output");
        self.tree.clear();
        rendered
    }
}

fn key_to_parts(key: Key) -> (Vec<String>, String) {
    let (name, labels) = key.into_parts();
    let mut parts = name.split('.').map(ToOwned::to_owned).collect::<Vec<_>>();
    let name = parts.pop().expect("name didn't have a single part");

    let labels = labels
        .into_iter()
        .map(Label::into_parts)
        .map(|(k, v)| format!("{}=\"{}\"", k, v))
        .collect::<Vec<_>>()
        .join(",");
    let label = if labels.is_empty() {
        String::new()
    } else {
        format!("{{{}}}", labels)
    };

    let fname = format!("{}{}", name, label);

    (parts, fname)
}

fn hist_to_values(
    name: String,
    hist: Histogram<u64>,
    quantiles: &[Quantile],
) -> Vec<(String, u64)> {
    let mut values = Vec::new();

    values.push((format!("{} count", name), hist.len()));
    for quantile in quantiles {
        let value = hist.value_at_quantile(quantile.value());
        values.push((format!("{} {}", name, quantile.label()), value));
    }

    values
}