referrerpolicy=no-referrer-when-downgrade
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
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
// Copyright (C) Parity Technologies (UK) Ltd.
// This file is part of Polkadot.

// Polkadot is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.

// Polkadot is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
// GNU General Public License for more details.

// You should have received a copy of the GNU General Public License
// along with Polkadot.  If not, see <http://www.gnu.org/licenses/>.

//! Prometheus metrics related to the validation host.

use polkadot_node_core_pvf_common::prepare::MemoryStats;
use polkadot_node_metrics::metrics::{self, prometheus};
use polkadot_node_subsystem::messages::PvfExecKind;

/// Validation host metrics.
#[derive(Default, Clone)]
pub struct Metrics(Option<MetricsInner>);

impl Metrics {
	/// Returns a handle to submit prepare workers metrics.
	pub(crate) fn prepare_worker(&'_ self) -> WorkerRelatedMetrics<'_> {
		WorkerRelatedMetrics { metrics: self, flavor: WorkerFlavor::Prepare }
	}

	/// Returns a handle to submit execute workers metrics.
	pub(crate) fn execute_worker(&'_ self) -> WorkerRelatedMetrics<'_> {
		WorkerRelatedMetrics { metrics: self, flavor: WorkerFlavor::Execute }
	}

	/// When preparation pipeline had a new item enqueued.
	pub(crate) fn prepare_enqueued(&self) {
		if let Some(metrics) = &self.0 {
			metrics.prepare_enqueued.inc();
		}
	}

	/// When preparation pipeline concluded working on an item.
	pub(crate) fn prepare_concluded(&self) {
		if let Some(metrics) = &self.0 {
			metrics.prepare_concluded.inc();
		}
	}

	/// When execution pipeline had a new item enqueued.
	pub(crate) fn execute_enqueued(&self) {
		if let Some(metrics) = &self.0 {
			metrics.execute_enqueued.inc();
		}
	}

	/// When execution pipeline finished executing a request.
	pub(crate) fn execute_finished(&self) {
		if let Some(metrics) = &self.0 {
			metrics.execute_finished.inc();
		}
	}

	/// Time between sending preparation request to a worker to having the response.
	pub(crate) fn time_preparation(
		&self,
	) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
		self.0.as_ref().map(|metrics| metrics.preparation_time.start_timer())
	}

	/// Time between sending execution request to a worker to having the response.
	pub(crate) fn time_execution(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
		self.0.as_ref().map(|metrics| metrics.execution_time.start_timer())
	}

	pub(crate) fn observe_execution_queued_time(&self, queued_for_millis: u32) {
		self.0.as_ref().map(|metrics| {
			metrics.execution_queued_time.observe(queued_for_millis as f64 / 1000 as f64)
		});
	}

	/// Observe memory stats for preparation.
	#[allow(unused_variables)]
	pub(crate) fn observe_preparation_memory_metrics(&self, memory_stats: MemoryStats) {
		if let Some(metrics) = &self.0 {
			#[cfg(target_os = "linux")]
			if let Some(max_rss) = memory_stats.max_rss {
				metrics.preparation_max_rss.observe(max_rss as f64);
			}

			#[cfg(any(target_os = "linux", feature = "jemalloc-allocator"))]
			if let Some(tracker_stats) = memory_stats.memory_tracker_stats {
				// We convert these stats from B to KB to match the unit of `ru_maxrss` from
				// `getrusage`.
				let max_resident_kb = (tracker_stats.resident / 1024) as f64;
				let max_allocated_kb = (tracker_stats.allocated / 1024) as f64;

				metrics.preparation_max_resident.observe(max_resident_kb);
				metrics.preparation_max_allocated.observe(max_allocated_kb);
			}

			metrics
				.preparation_peak_tracked_allocation
				.observe((memory_stats.peak_tracked_alloc / 1024) as f64);
		}
	}

	pub(crate) fn observe_code_size(&self, code_size: usize) {
		if let Some(metrics) = &self.0 {
			metrics.code_size.observe(code_size as f64);
		}
	}

	pub(crate) fn observe_pov_size(&self, pov_size: usize, compressed: bool) {
		if let Some(metrics) = &self.0 {
			metrics
				.pov_size
				.with_label_values(&[if compressed { "true" } else { "false" }])
				.observe(pov_size as f64);
		}
	}

	/// When preparation pipeline concluded working on an item.
	pub(crate) fn on_execute_kind(&self, kind: PvfExecKind) {
		if let Some(metrics) = &self.0 {
			metrics.exec_kind_selected.with_label_values(&[kind.as_str()]).inc();
		}
	}
}

#[derive(Clone)]
struct MetricsInner {
	worker_spawning: prometheus::CounterVec<prometheus::U64>,
	worker_spawned: prometheus::CounterVec<prometheus::U64>,
	worker_retired: prometheus::CounterVec<prometheus::U64>,
	prepare_enqueued: prometheus::Counter<prometheus::U64>,
	prepare_concluded: prometheus::Counter<prometheus::U64>,
	execute_enqueued: prometheus::Counter<prometheus::U64>,
	execute_finished: prometheus::Counter<prometheus::U64>,
	preparation_time: prometheus::Histogram,
	execution_time: prometheus::Histogram,
	execution_queued_time: prometheus::Histogram,
	#[cfg(target_os = "linux")]
	preparation_max_rss: prometheus::Histogram,
	// Max. allocated memory, tracked by Jemallocator, polling-based
	#[cfg(any(target_os = "linux", feature = "jemalloc-allocator"))]
	preparation_max_allocated: prometheus::Histogram,
	// Max. resident memory, tracked by Jemallocator, polling-based
	#[cfg(any(target_os = "linux", feature = "jemalloc-allocator"))]
	preparation_max_resident: prometheus::Histogram,
	// Peak allocation value, tracked by tracking-allocator
	preparation_peak_tracked_allocation: prometheus::Histogram,
	pov_size: prometheus::HistogramVec,
	code_size: prometheus::Histogram,
	exec_kind_selected: prometheus::CounterVec<prometheus::U64>,
}

impl metrics::Metrics for Metrics {
	fn try_register(registry: &prometheus::Registry) -> Result<Self, prometheus::PrometheusError> {
		let inner = MetricsInner {
			worker_spawning: prometheus::register(
				prometheus::CounterVec::new(
					prometheus::Opts::new(
						"polkadot_pvf_worker_spawning",
						"The total number of workers began to spawn",
					),
					&["flavor"],
				)?,
				registry,
			)?,
			worker_spawned: prometheus::register(
				prometheus::CounterVec::new(
					prometheus::Opts::new(
						"polkadot_pvf_worker_spawned",
						"The total number of workers spawned successfully",
					),
					&["flavor"],
				)?,
				registry,
			)?,
			worker_retired: prometheus::register(
				prometheus::CounterVec::new(
					prometheus::Opts::new(
						"polkadot_pvf_worker_retired",
						"The total number of workers retired, either killed by the host or died on duty",
					),
					&["flavor"],
				)?,
				registry,
			)?,
			prepare_enqueued: prometheus::register(
				prometheus::Counter::new(
					"polkadot_pvf_prepare_enqueued",
					"The total number of jobs enqueued into the preparation pipeline"
				)?,
				registry,
			)?,
			prepare_concluded: prometheus::register(
				prometheus::Counter::new(
					"polkadot_pvf_prepare_concluded",
					"The total number of jobs concluded in the preparation pipeline"
				)?,
				registry,
			)?,
			execute_enqueued: prometheus::register(
				prometheus::Counter::new(
					"polkadot_pvf_execute_enqueued",
					"The total number of jobs enqueued into the execution pipeline"
				)?,
				registry,
			)?,
			execute_finished: prometheus::register(
				prometheus::Counter::new(
					"polkadot_pvf_execute_finished",
					"The total number of jobs done in the execution pipeline"
				)?,
				registry,
			)?,
			preparation_time: prometheus::register(
				prometheus::Histogram::with_opts(
					prometheus::HistogramOpts::new(
						"polkadot_pvf_preparation_time",
						"Time spent in preparing PVF artifacts in seconds",
					)
					.buckets(vec![
						// This is synchronized with the `DEFAULT_PRECHECK_PREPARATION_TIMEOUT=60s`
						// and `DEFAULT_LENIENT_PREPARATION_TIMEOUT=360s` constants found in
						// node/core/candidate-validation/src/lib.rs
						0.1,
						0.5,
						1.0,
						2.0,
						3.0,
						10.0,
						20.0,
						30.0,
						60.0,
						120.0,
						240.0,
						360.0,
						480.0,
					]),
				)?,
				registry,
			)?,
			execution_time: prometheus::register(
				prometheus::Histogram::with_opts(
					prometheus::HistogramOpts::new(
						"polkadot_pvf_execution_time",
						"Time spent in executing PVFs",
					).buckets(vec![
						// This is synchronized with `DEFAULT_APPROVAL_EXECUTION_TIMEOUT` and
						// `DEFAULT_BACKING_EXECUTION_TIMEOUT` constants in
						// node/core/candidate-validation/src/lib.rs
						0.01,
						0.025,
						0.05,
						0.1,
						0.25,
						0.5,
						1.0,
						2.0,
						3.0,
						4.0,
						5.0,
						6.0,
						8.0,
						10.0,
						12.0,
					]),
				)?,
				registry,
			)?,
			execution_queued_time: prometheus::register(
				prometheus::Histogram::with_opts(
					prometheus::HistogramOpts::new(
						"polkadot_pvf_execution_queued_time",
						"Time spent in queue waiting for PVFs execution job to be assigned",
					).buckets(vec![
						0.01,
						0.025,
						0.05,
						0.1,
						0.25,
						0.5,
						1.0,
						2.0,
						3.0,
						4.0,
						5.0,
						6.0,
						12.0,
						24.0,
						48.0,
					]),
				)?,
				registry,
			)?,
			#[cfg(target_os = "linux")]
			preparation_max_rss: prometheus::register(
				prometheus::Histogram::with_opts(
					prometheus::HistogramOpts::new(
						"polkadot_pvf_preparation_max_rss",
						"ru_maxrss (maximum resident set size) observed for preparation (in kilobytes)",
					).buckets(
						prometheus::exponential_buckets(8192.0, 2.0, 10)
							.expect("arguments are always valid; qed"),
					),
				)?,
				registry,
			)?,
			#[cfg(any(target_os = "linux", feature = "jemalloc-allocator"))]
			preparation_max_resident: prometheus::register(
				prometheus::Histogram::with_opts(
					prometheus::HistogramOpts::new(
						"polkadot_pvf_preparation_max_resident",
						"max resident memory observed for preparation (in kilobytes)",
					).buckets(
						prometheus::exponential_buckets(8192.0, 2.0, 10)
							.expect("arguments are always valid; qed"),
					),
				)?,
				registry,
			)?,
			#[cfg(any(target_os = "linux", feature = "jemalloc-allocator"))]
			preparation_max_allocated: prometheus::register(
				prometheus::Histogram::with_opts(
					prometheus::HistogramOpts::new(
						"polkadot_pvf_preparation_max_allocated",
						"max allocated memory observed for preparation (in kilobytes)",
					).buckets(
						prometheus::exponential_buckets(8192.0, 2.0, 10)
							.expect("arguments are always valid; qed"),
					),
				)?,
				registry,
			)?,
			preparation_peak_tracked_allocation: prometheus::register(
				prometheus::Histogram::with_opts(
					prometheus::HistogramOpts::new(
						"polkadot_pvf_preparation_peak_tracked_allocation",
						"peak allocation observed for preparation (in kilobytes)",
					).buckets(
						prometheus::exponential_buckets(8192.0, 2.0, 10)
							.expect("arguments are always valid; qed"),
					),
				)?,
				registry,
			)?,
			// The following metrics was moved here from the candidate valiidation subsystem.
			// Names are kept to avoid breaking dashboards and stuff.
			pov_size: prometheus::register(
				prometheus::HistogramVec::new(
					prometheus::HistogramOpts::new(
						"polkadot_parachain_candidate_validation_pov_size",
						"The compressed and decompressed size of the proof of validity of a candidate",
					)
					.buckets(
						prometheus::exponential_buckets(16384.0, 2.0, 10)
							.expect("arguments are always valid; qed"),
					),
					&["compressed"],
				)?,
				registry,
			)?,
			code_size: prometheus::register(
				prometheus::Histogram::with_opts(
					prometheus::HistogramOpts::new(
						"polkadot_parachain_candidate_validation_code_size",
						"The size of the decompressed WASM validation blob used for checking a candidate",
					)
					.buckets(
						prometheus::exponential_buckets(16384.0, 2.0, 10)
							.expect("arguments are always valid; qed"),
					),
				)?,
				registry,
			)?,
			exec_kind_selected: prometheus::register(
				prometheus::CounterVec::new(
					prometheus::Opts::new(
						"polkadot_pvf_exec_kind_selected",
						"The total number of selected execute kinds",
					),
					&["priority"],
				)?,
				registry,
			)?,
		};
		Ok(Metrics(Some(inner)))
	}
}

enum WorkerFlavor {
	Prepare,
	Execute,
}

impl WorkerFlavor {
	fn as_label(&self) -> &'static str {
		match *self {
			WorkerFlavor::Prepare => "prepare",
			WorkerFlavor::Execute => "execute",
		}
	}
}

pub(crate) struct WorkerRelatedMetrics<'a> {
	metrics: &'a Metrics,
	flavor: WorkerFlavor,
}

impl<'a> WorkerRelatedMetrics<'a> {
	/// When the spawning of a worker started.
	pub(crate) fn on_begin_spawn(&self) {
		if let Some(metrics) = &self.metrics.0 {
			metrics.worker_spawning.with_label_values(&[self.flavor.as_label()]).inc();
		}
	}

	/// When the worker successfully spawned.
	pub(crate) fn on_spawned(&self) {
		if let Some(metrics) = &self.metrics.0 {
			metrics.worker_spawned.with_label_values(&[self.flavor.as_label()]).inc();
		}
	}

	/// When the worker was killed or died.
	pub(crate) fn on_retired(&self) {
		if let Some(metrics) = &self.metrics.0 {
			metrics.worker_retired.with_label_values(&[self.flavor.as_label()]).inc();
		}
	}
}