Add metrics for out of view statement distribution errors (#4972)

* Add a simple metric for statements out-of-view * Avoid repeated out-of-view peer reputation change messages * Log reporting status * Address review comments * Use counter to store a number of unexpected messages from a peer * Distinguish different unexpected statements in the metrics * Fix labels cardinality * Rename metric name to `statements_unexpected` * Move metrics to a separate unit, avoid unnecessary enum * Prefer specific methods in lieu of public constants
2026-06-19 00:41:03 +00:00 · 2022-02-25 20:29:51 +00:00
parent 86f2d65a72
commit 44f66825c7
3 changed files with 218 additions and 115 deletions
@@ -0,0 +1,165 @@
+// Copyright 2022 Parity Technologies (UK) Ltd.
+// 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/>.
+
+//! Metrics for the statement distribution module
+
+use polkadot_node_subsystem_util::metrics::{self, prometheus};
+
+#[derive(Clone)]
+struct MetricsInner {
+	statements_distributed: prometheus::Counter<prometheus::U64>,
+	sent_requests: prometheus::Counter<prometheus::U64>,
+	received_responses: prometheus::CounterVec<prometheus::U64>,
+	active_leaves_update: prometheus::Histogram,
+	share: prometheus::Histogram,
+	network_bridge_update_v1: prometheus::Histogram,
+	statements_unexpected: prometheus::CounterVec<prometheus::U64>,
+}
+
+/// Statement Distribution metrics.
+#[derive(Default, Clone)]
+pub struct Metrics(Option<MetricsInner>);
+
+impl Metrics {
+	/// Update statements distributed counter
+	pub fn on_statement_distributed(&self) {
+		if let Some(metrics) = &self.0 {
+			metrics.statements_distributed.inc();
+		}
+	}
+
+	/// Update sent requests counter
+	/// This counter is updated merely for the statements sent via request/response method,
+	/// meaning that it counts large statements only
+	pub fn on_sent_request(&self) {
+		if let Some(metrics) = &self.0 {
+			metrics.sent_requests.inc();
+		}
+	}
+
+	/// Update counters for the received responses with `succeeded` or `failed` labels
+	/// These counters are updated merely for the statements received via request/response method,
+	/// meaning that they count large statements only
+	pub fn on_received_response(&self, success: bool) {
+		if let Some(metrics) = &self.0 {
+			let label = if success { "succeeded" } else { "failed" };
+			metrics.received_responses.with_label_values(&[label]).inc();
+		}
+	}
+
+	/// Provide a timer for `active_leaves_update` which observes on drop.
+	pub fn time_active_leaves_update(
+		&self,
+	) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.active_leaves_update.start_timer())
+	}
+
+	/// Provide a timer for `share` which observes on drop.
+	pub fn time_share(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.share.start_timer())
+	}
+
+	/// Provide a timer for `network_bridge_update_v1` which observes on drop.
+	pub fn time_network_bridge_update_v1(
+		&self,
+	) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.network_bridge_update_v1.start_timer())
+	}
+
+	/// Update the out-of-view statements counter for unexpected valid statements
+	pub fn on_unexpected_statement_valid(&self) {
+		if let Some(metrics) = &self.0 {
+			metrics.statements_unexpected.with_label_values(&["valid"]).inc();
+		}
+	}
+
+	/// Update the out-of-view statements counter for unexpected seconded statements
+	pub fn on_unexpected_statement_seconded(&self) {
+		if let Some(metrics) = &self.0 {
+			metrics.statements_unexpected.with_label_values(&["seconded"]).inc();
+		}
+	}
+
+	/// Update the out-of-view statements counter for unexpected large statements
+	pub fn on_unexpected_statement_large(&self) {
+		if let Some(metrics) = &self.0 {
+			metrics.statements_unexpected.with_label_values(&["large"]).inc();
+		}
+	}
+}
+
+impl metrics::Metrics for Metrics {
+	fn try_register(
+		registry: &prometheus::Registry,
+	) -> std::result::Result<Self, prometheus::PrometheusError> {
+		let metrics = MetricsInner {
+			statements_distributed: prometheus::register(
+				prometheus::Counter::new(
+					"polkadot_parachain_statements_distributed_total",
+					"Number of candidate validity statements distributed to other peers.",
+				)?,
+				registry,
+			)?,
+			sent_requests: prometheus::register(
+				prometheus::Counter::new(
+					"polkadot_parachain_statement_distribution_sent_requests_total",
+					"Number of large statement fetching requests sent.",
+				)?,
+				registry,
+			)?,
+			received_responses: prometheus::register(
+				prometheus::CounterVec::new(
+					prometheus::Opts::new(
+						"polkadot_parachain_statement_distribution_received_responses_total",
+						"Number of received responses for large statement data.",
+					),
+					&["success"],
+				)?,
+				registry,
+			)?,
+			active_leaves_update: prometheus::register(
+				prometheus::Histogram::with_opts(prometheus::HistogramOpts::new(
+					"polkadot_parachain_statement_distribution_active_leaves_update",
+					"Time spent within `statement_distribution::active_leaves_update`",
+				))?,
+				registry,
+			)?,
+			share: prometheus::register(
+				prometheus::Histogram::with_opts(prometheus::HistogramOpts::new(
+					"polkadot_parachain_statement_distribution_share",
+					"Time spent within `statement_distribution::share`",
+				))?,
+				registry,
+			)?,
+			network_bridge_update_v1: prometheus::register(
+				prometheus::Histogram::with_opts(prometheus::HistogramOpts::new(
+					"polkadot_parachain_statement_distribution_network_bridge_update_v1",
+					"Time spent within `statement_distribution::network_bridge_update_v1`",
+				))?,
+				registry,
+			)?,
+			statements_unexpected: prometheus::register(
+				prometheus::CounterVec::new(
+					prometheus::Opts::new(
+						"polkadot_parachain_statement_distribution_statements_unexpected",
+						"Number of statements that were not expected to be received.",
+					),
+					&["type"],
+				)?,
+				registry,
+			)?,
+		};
+		Ok(Metrics(Some(metrics)))
+	}
+}