File size: 9,435 Bytes
d5bfab8 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 |
use super::{MetricEvent, MetricsPrometheus, MovingAverage, Recorder, CoordinatorWorkerMessage};
use std::time::{Duration, Instant};
use prometheus_client::encoding::text::encode;
use prometheus_client::registry::Registry;
use std::sync::{Arc, Mutex};
use bastion::prelude::*;
pub struct MetricsWorker;
impl MetricsWorker {
/// Print metrics every second to commandline.
pub fn start_without_server() -> anyhow::Result<()> {
Bastion::supervisor(|supervisor| {
supervisor.children(|children| {
children
.with_redundancy(1)
.with_distributor(Distributor::named("metrics_worker"))
.with_exec(metrics_worker_print)
})
}).expect("Unable to create metrics_worker_print");
Ok(())
}
/// Runs a webserver with realtime metrics, so bottlenecks can be identified.
pub fn start_with_server(listen_on_port: u16, number_of_workers: u64) -> anyhow::Result<()> {
println!("miner metrics can be downloaded here: http://localhost:{}/metrics", listen_on_port);
let mut registry = <Registry>::default();
let metrics = MetricsPrometheus::new(&mut registry);
metrics.number_of_workers.set(number_of_workers);
let registry2: MyRegistry = Arc::new(Mutex::new(registry));
let _ = tokio::spawn(async move {
let result = webserver_with_metrics(registry2, listen_on_port).await;
if let Err(error) = result {
error!("webserver thread failed with error: {:?}", error);
}
});
Bastion::supervisor(|supervisor| {
supervisor.children(|children| {
children
.with_redundancy(1)
.with_distributor(Distributor::named("metrics_worker"))
.with_exec(move |ctx: BastionContext| {
let metrics_clone = metrics.clone();
async move {
metrics_worker_server(
ctx,
metrics_clone,
).await
}
})
})
}).expect("Unable to create metrics_worker_webserver");
Ok(())
}
}
#[derive(Clone)]
struct State {
registry: MyRegistry,
}
type MyRegistryInner = prometheus_client::registry::Registry<std::boxed::Box<dyn prometheus_client::encoding::text::SendSyncEncodeMetric>>;
type MyRegistry = std::sync::Arc<std::sync::Mutex<MyRegistryInner>>;
async fn webserver_with_metrics(registry: MyRegistry, listen_port: u16) -> std::result::Result<(), Box<dyn std::error::Error>> {
let mut app = tide::with_state(State {
registry: registry,
});
app.at("/").get(|_| async { Ok("Hello, world!") });
app.at("/sync").get(|_| async {
let distributor = Distributor::named("coordinator_worker");
let tell_result = distributor.tell_everyone(CoordinatorWorkerMessage::TriggerSync);
if let Err(error) = tell_result {
let response = tide::Response::builder(500)
.body(format!("webserver_with_metrics: /sync - Unable to send TriggerSync to coordinator_worker_distributor. {:?}", error))
.content_type("text/plain; charset=utf-8")
.build();
return Ok(response);
}
let response = tide::Response::builder(200)
.body("did send TriggerSync")
.content_type("text/plain; charset=utf-8")
.build();
Ok(response)
});
app.at("/metrics")
.get(|req: tide::Request<State>| async move {
let mut encoded = Vec::new();
{
let registry = match req.state().registry.lock() {
Ok(value) => value,
Err(error) => {
error!("webserver_with_metrics: /metrics - Unable to lock registry. {:?}", error);
let response = tide::Response::builder(500)
.body(format!("webserver_with_metrics: /metrics - Unable to lock registry. {:?}", error))
.content_type("text/plain; charset=utf-8")
.build();
return Ok(response);
}
};
match encode(&mut encoded, ®istry) {
Ok(()) => {},
Err(error) => {
error!("webserver_with_metrics: /metrics - Unable to encode registry. {:?}", error);
let response = tide::Response::builder(500)
.body(format!("webserver_with_metrics: /metrics - Unable to encode registry. {:?}", error))
.content_type("text/plain; charset=utf-8")
.build();
return Ok(response);
}
}
}
let response = tide::Response::builder(200)
.body(encoded)
.content_type("application/openmetrics-text; version=1.0.0; charset=utf-8")
.build();
Ok(response)
});
let server_address = format!("localhost:{}", listen_port);
app.listen(server_address).await?;
Ok(())
}
/// Inspect metrics via the Graphana dashboard.
///
/// Underneeth Graphana accesses the data via Prometheus.
///
/// This function forwards metrics events to Prometheus.
async fn metrics_worker_server(ctx: BastionContext, metrics: MetricsPrometheus) -> Result<(), ()> {
debug!("metrics_worker_server is ready");
let mut progress_time = Instant::now();
let mut miner_iteration_count: u64 = 0;
let mut moving_average = MovingAverage::new();
loop {
let elapsed: u128 = progress_time.elapsed().as_millis();
if elapsed >= 1000 {
// Compute average number of iterations over the last second.
moving_average.insert(miner_iteration_count);
let weighted_average: u64 = moving_average.average();
moving_average.rotate();
metrics.number_of_iteration_now.set(weighted_average);
progress_time = Instant::now();
miner_iteration_count = 0;
}
let timeout = Duration::from_millis(1000);
let message: SignedMessage = match ctx.try_recv_timeout(timeout).await {
Ok(message) => message,
Err(error) => {
if let ReceiveError::Timeout(_duration) = error {
debug!("metrics_worker_server: timeout happened");
continue;
}
error!("metrics_worker_server: Unknown error happened. error: {:?}", error);
continue;
}
};
MessageHandler::new(message)
.on_tell(|metric_event: MetricEvent, _| {
if let MetricEvent::General { number_of_iterations, .. } = metric_event {
miner_iteration_count += number_of_iterations;
}
metrics.record(&metric_event);
})
.on_fallback(|unknown, _sender_addr| {
error!(
"metrics_worker_server {}, received an unknown message!:\n{:?}",
ctx.current().id(),
unknown
);
});
}
}
/// Inspect metrics from commandline.
///
/// Prints stats with 1 second interval.
async fn metrics_worker_print(ctx: BastionContext) -> Result<(), ()> {
debug!("metrics_worker_print is ready");
let mut progress_time = Instant::now();
let mut miner_iteration_count: u64 = 0;
let mut metric_event_count: u64 = 0;
loop {
let elapsed: u128 = progress_time.elapsed().as_millis();
if elapsed >= 1000 {
if miner_iteration_count > 0 {
println!("miner_iterations: {}", miner_iteration_count);
} else {
debug!("metrics_worker_print: metric_events: {} miner_iterations: {} - no activity", metric_event_count, miner_iteration_count);
}
progress_time = Instant::now();
miner_iteration_count = 0;
metric_event_count = 0;
}
let timeout = Duration::from_millis(1000);
let message: SignedMessage = match ctx.try_recv_timeout(timeout).await {
Ok(message) => message,
Err(error) => {
if let ReceiveError::Timeout(_duration) = error {
debug!("metrics_worker_print: timeout happened");
continue;
}
error!("metrics_worker_print: Unknown error happened. error: {:?}", error);
continue;
}
};
MessageHandler::new(message)
.on_tell(|metric_event: MetricEvent, _| {
metric_event_count += 1;
if let MetricEvent::General { number_of_iterations, .. } = metric_event {
miner_iteration_count += number_of_iterations;
}
})
.on_fallback(|unknown, _sender_addr| {
error!(
"metrics_worker_print {}, received an unknown message!:\n{:?}",
ctx.current().id(),
unknown
);
});
}
}
|