Add Prometheus timers to the subsystems (#1923)

* reexport prometheus-super for ease of use of other subsystems

* add some prometheus timers for collation generation subsystem

* add timing metrics to av-store

* add metrics to candidate backing

* add timing metric to bitfield signing

* add timing metrics to candidate selection

* add timing metrics to candidate-validation

* add timing metrics to chain-api

* add timing metrics to provisioner

* add timing metrics to runtime-api

* add timing metrics to availability-distribution

* add timing metrics to bitfield-distribution

* add timing metrics to collator protocol: collator side

* add timing metrics to collator protocol: validator side

* fix candidate validation test failures

* add timing metrics to pov distribution

* add timing metrics to statement-distribution

* use substrate_prometheus_endpoint prometheus reexport instead of prometheus_super

* don't include JOB_DELAY in bitfield-signing metrics

* give adder-collator ability to easily export its genesis-state and validation code

* wip: adder-collator pushbutton script

* don't attempt to register the adder-collator automatically

Instead, get these values with

```sh
target/release/adder-collator export-genesis-state
target/release/adder-collator export-genesis-wasm
```

And then register the parachain on https://polkadot.js.org/apps/?rpc=ws%3A%2F%2F127.0.0.1%3A9944#/explorer

To collect prometheus data, after running the script, create `prometheus.yml` per the instructions
at https://www.notion.so/paritytechnologies/Setting-up-Prometheus-locally-835cb3a9df7541a781c381006252b5ff
and then run:

```sh
docker run -v `pwd`/prometheus.yml:/etc/prometheus/prometheus.yml:z --network host prom/prometheus
```

Demonstrates that data makes it across to prometheus, though it is likely to be useful in the future
to tweak the buckets.

* Update parachain/test-parachains/adder/collator/src/cli.rs

Co-authored-by: Andronik Ordian <write@reusable.software>

* use the grandpa-pause parameter

* skip metrics in tracing instrumentation

* remove unnecessary grandpa_pause cli param

Co-authored-by: Andronik Ordian <write@reusable.software>

polkadot/node/collation-generation/src/lib.rs

@@ -189,7 +189,11 @@ async fn handle_new_activations<Context: SubsystemContext>(
 	// follow the procedure from the guide:
 	// https://w3f.github.io/parachain-implementers-guide/node/collators/collation-generation.html
 
+	let _overall_timer = metrics.time_new_activations();
+
 	for relay_parent in activated.iter().copied() {
+		let _relay_parent_timer = metrics.time_new_activations_relay_parent();
+
 		// double-future magic happens here: the first layer of requests takes a mutable borrow of the context, and
 		// returns a receiver. The second layer of requests actually polls those receivers to completion.
 		let (availability_cores, validators) = join!(
@@ -201,6 +205,8 @@ async fn handle_new_activations<Context: SubsystemContext>(
 		let n_validators = validators??.len();
 
 		for core in availability_cores {
+			let _availability_core_timer = metrics.time_new_activations_availability_core();
+
 			let (scheduled_core, assumption) = match core {
 				CoreState::Scheduled(scheduled_core) => {
 					(scheduled_core, OccupiedCoreAssumption::Free)
@@ -335,6 +341,9 @@ fn erasure_root(
 #[derive(Clone)]
 struct MetricsInner {
 	collations_generated_total: prometheus::Counter<prometheus::U64>,
+	new_activations_overall: prometheus::Histogram,
+	new_activations_per_relay_parent: prometheus::Histogram,
+	new_activations_per_availability_core: prometheus::Histogram,
 }
 
 /// CollationGenerationSubsystem metrics.
@@ -347,6 +356,21 @@ impl Metrics {
 			metrics.collations_generated_total.inc();
 		}
 	}
+
+	/// Provide a timer for new activations which updates on drop.
+	fn time_new_activations(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.new_activations_overall.start_timer())
+	}
+
+	/// Provide a timer per relay parents which updates on drop.
+	fn time_new_activations_relay_parent(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.new_activations_per_relay_parent.start_timer())
+	}
+
+	/// Provide a timer per availability core which updates on drop.
+	fn time_new_activations_availability_core(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.new_activations_per_availability_core.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -359,6 +383,33 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			new_activations_overall: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_collation_generation_new_activations",
+						"Time spent within fn handle_new_activations",
+					)
+				)?,
+				registry,
+			)?,
+			new_activations_per_relay_parent: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_collation_generation_per_relay_parent",
+						"Time spent handling a particular relay parent within fn handle_new_activations"
+					)
+				)?,
+				registry,
+			)?,
+			new_activations_per_availability_core: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_collation_generation_per_availability_core",
+						"Time spent handling a particular availability core for a relay parent in fn handle_new_activations",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}

polkadot/node/core/av-store/src/lib.rs

@@ -313,6 +313,8 @@ impl AvailabilityStoreSubsystem {
 	// Perform pruning of PoVs
 	#[tracing::instrument(level = "trace", skip(self), fields(subsystem = LOG_TARGET))]
 	fn prune_povs(&self) -> Result<(), Error> {
+		let _timer = self.metrics.time_prune_povs();
+
 		let mut tx = DBTransaction::new();
 		let mut pov_pruning = pov_pruning(&self.inner).unwrap_or_default();
 		let now = PruningDelay::now()?;
@@ -338,6 +340,8 @@ impl AvailabilityStoreSubsystem {
 	// Perform pruning of chunks.
 	#[tracing::instrument(level = "trace", skip(self), fields(subsystem = LOG_TARGET))]
 	fn prune_chunks(&self) -> Result<(), Error> {
+		let _timer = self.metrics.time_prune_chunks();
+
 		let mut tx = DBTransaction::new();
 		let mut chunk_pruning = chunk_pruning(&self.inner).unwrap_or_default();
 		let now = PruningDelay::now()?;
@@ -522,7 +526,7 @@ where
 					ActiveLeavesUpdate { activated, .. })
 				) => {
 					for activated in activated.into_iter() {
-						process_block_activated(ctx, &subsystem.inner, activated).await?;
+						process_block_activated(ctx, &subsystem.inner, activated, &subsystem.metrics).await?;
 					}
 				}
 				FromOverseer::Signal(OverseerSignal::BlockFinalized(hash)) => {
@@ -561,6 +565,8 @@ async fn process_block_finalized<Context>(
 where
 	Context: SubsystemContext<Message=AvailabilityStoreMessage>
 {
+	let _timer = subsystem.metrics.time_process_block_finalized();
+
 	let block_number = get_block_number(ctx, hash).await?;
 
 	if let Some(mut pov_pruning) = pov_pruning(db) {
@@ -606,15 +612,18 @@ where
 	Ok(())
 }
 
-#[tracing::instrument(level = "trace", skip(ctx, db), fields(subsystem = LOG_TARGET))]
+#[tracing::instrument(level = "trace", skip(ctx, db, metrics), fields(subsystem = LOG_TARGET))]
 async fn process_block_activated<Context>(
 	ctx: &mut Context,
 	db: &Arc<dyn KeyValueDB>,
 	hash: Hash,
+	metrics: &Metrics,
 ) -> Result<(), Error>
 where
 	Context: SubsystemContext<Message=AvailabilityStoreMessage>
 {
+	let _timer = metrics.time_block_activated();
+
 	let events = match request_candidate_events(ctx, hash).await {
 		Ok(events) => events,
 		Err(err) => {
@@ -697,6 +706,8 @@ where
 {
 	use AvailabilityStoreMessage::*;
 
+	let _timer = subsystem.metrics.time_process_message();
+
 	match msg {
 		QueryAvailableData(hash, tx) => {
 			tx.send(available_data(&subsystem.inner, &hash).map(|d| d.data))
@@ -860,6 +871,8 @@ fn store_available_data(
 	n_validators: u32,
 	available_data: AvailableData,
 ) -> Result<(), Error> {
+	let _timer = subsystem.metrics.time_store_available_data();
+
 	let mut tx = DBTransaction::new();
 
 	let block_number = available_data.validation_data.block_number;
@@ -927,6 +940,8 @@ fn store_chunk(
 	chunk: ErasureChunk,
 	block_number: BlockNumber,
 ) -> Result<(), Error> {
+	let _timer = subsystem.metrics.time_store_chunk();
+
 	let mut tx = DBTransaction::new();
 
 	let dbkey = erasure_chunk_key(candidate_hash, chunk.index);
@@ -977,6 +992,8 @@ fn get_chunk(
 	candidate_hash: &CandidateHash,
 	index: u32,
 ) -> Result<Option<ErasureChunk>, Error> {
+	let _timer = subsystem.metrics.time_get_chunk();
+
 	if let Some(chunk) = query_inner(
 		&subsystem.inner,
 		columns::DATA,
@@ -1059,6 +1076,14 @@ fn get_chunks(data: &AvailableData, n_validators: usize, metrics: &Metrics) -> R
 #[derive(Clone)]
 struct MetricsInner {
 	received_availability_chunks_total: prometheus::Counter<prometheus::U64>,
+	prune_povs: prometheus::Histogram,
+	prune_chunks: prometheus::Histogram,
+	process_block_finalized: prometheus::Histogram,
+	block_activated: prometheus::Histogram,
+	process_message: prometheus::Histogram,
+	store_available_data: prometheus::Histogram,
+	store_chunk: prometheus::Histogram,
+	get_chunk: prometheus::Histogram,
 }
 
 /// Availability metrics.
@@ -1074,6 +1099,46 @@ impl Metrics {
 			metrics.received_availability_chunks_total.inc_by(by);
 		}
 	}
+
+	/// Provide a timer for `prune_povs` which observes on drop.
+	fn time_prune_povs(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.prune_povs.start_timer())
+	}
+
+	/// Provide a timer for `prune_chunks` which observes on drop.
+	fn time_prune_chunks(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.prune_chunks.start_timer())
+	}
+
+	/// Provide a timer for `process_block_finalized` which observes on drop.
+	fn time_process_block_finalized(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.process_block_finalized.start_timer())
+	}
+
+	/// Provide a timer for `block_activated` which observes on drop.
+	fn time_block_activated(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.block_activated.start_timer())
+	}
+
+	/// Provide a timer for `process_message` which observes on drop.
+	fn time_process_message(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.process_message.start_timer())
+	}
+
+	/// Provide a timer for `store_available_data` which observes on drop.
+	fn time_store_available_data(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.store_available_data.start_timer())
+	}
+
+	/// Provide a timer for `store_chunk` which observes on drop.
+	fn time_store_chunk(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.store_chunk.start_timer())
+	}
+
+	/// Provide a timer for `get_chunk` which observes on drop.
+	fn time_get_chunk(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.get_chunk.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -1086,6 +1151,78 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			prune_povs: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_av_store_prune_povs",
+						"Time spent within `av_store::prune_povs`",
+					)
+				)?,
+				registry,
+			)?,
+			prune_chunks: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_av_store_prune_chunks",
+						"Time spent within `av_store::prune_chunks`",
+					)
+				)?,
+				registry,
+			)?,
+			process_block_finalized: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_av_store_process_block_finalized",
+						"Time spent within `av_store::block_finalized`",
+					)
+				)?,
+				registry,
+			)?,
+			block_activated: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_av_store_block_activated",
+						"Time spent within `av_store::block_activated`",
+					)
+				)?,
+				registry,
+			)?,
+			process_message: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_av_store_process_message",
+						"Time spent within `av_store::process_message`",
+					)
+				)?,
+				registry,
+			)?,
+			store_available_data: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_av_store_store_available_data",
+						"Time spent within `av_store::store_available_data`",
+					)
+				)?,
+				registry,
+			)?,
+			store_chunk: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_av_store_store_chunk",
+						"Time spent within `av_store::store_chunk`",
+					)
+				)?,
+				registry,
+			)?,
+			get_chunk: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_av_store_get_chunk",
+						"Time spent within `av_store::get_chunk`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}

polkadot/node/core/backing/src/lib.rs

@@ -481,8 +481,11 @@ impl CandidateBackingJob {
 
 	#[tracing::instrument(level = "trace", skip(self), fields(subsystem = LOG_TARGET))]
 	async fn process_msg(&mut self, msg: CandidateBackingMessage) -> Result<(), Error> {
+
 		match msg {
 			CandidateBackingMessage::Second(_, candidate, pov) => {
+				let _timer = self.metrics.time_process_second();
+
 				// Sanity check that candidate is from our assignment.
 				if candidate.descriptor().para_id != self.assignment {
 					return Ok(());
@@ -509,6 +512,8 @@ impl CandidateBackingJob {
 				}
 			}
 			CandidateBackingMessage::Statement(_, statement) => {
+				let _timer = self.metrics.time_process_statement();
+
 				self.check_statement_signature(&statement)?;
 				match self.maybe_validate_and_import(statement).await {
 					Err(Error::ValidationFailed(_)) => return Ok(()),
@@ -517,6 +522,8 @@ impl CandidateBackingJob {
 				}
 			}
 			CandidateBackingMessage::GetBackedCandidates(_, tx) => {
+				let _timer = self.metrics.time_get_backed_candidates();
+
 				let backed = self.get_backed();
 
 				tx.send(backed).map_err(|data| Error::Send(data))?;
@@ -898,7 +905,10 @@ impl util::JobTrait for CandidateBackingJob {
 #[derive(Clone)]
 struct MetricsInner {
 	signed_statements_total: prometheus::Counter<prometheus::U64>,
-	candidates_seconded_total: prometheus::Counter<prometheus::U64>
+	candidates_seconded_total: prometheus::Counter<prometheus::U64>,
+	process_second: prometheus::Histogram,
+	process_statement: prometheus::Histogram,
+	get_backed_candidates: prometheus::Histogram,
 }
 
 /// Candidate backing metrics.
@@ -917,6 +927,21 @@ impl Metrics {
 			metrics.candidates_seconded_total.inc();
 		}
 	}
+
+	/// Provide a timer for handling `CandidateBackingMessage:Second` which observes on drop.
+	fn time_process_second(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.process_second.start_timer())
+	}
+
+	/// Provide a timer for handling `CandidateBackingMessage::Statement` which observes on drop.
+	fn time_process_statement(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.process_statement.start_timer())
+	}
+
+	/// Provide a timer for handling `CandidateBackingMessage::GetBackedCandidates` which observes on drop.
+	fn time_get_backed_candidates(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.get_backed_candidates.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -924,18 +949,45 @@ impl metrics::Metrics for Metrics {
 		let metrics = MetricsInner {
 			signed_statements_total: prometheus::register(
 				prometheus::Counter::new(
-					"parachain_signed_statements_total",
+					"parachain_candidate_backing_signed_statements_total",
 					"Number of statements signed.",
 				)?,
 				registry,
 			)?,
 			candidates_seconded_total: prometheus::register(
 				prometheus::Counter::new(
-					"parachain_candidates_seconded_total",
+					"parachain_candidate_backing_candidates_seconded_total",
 					"Number of candidates seconded.",
 				)?,
 				registry,
 			)?,
+			process_second: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_candidate_backing_process_second",
+						"Time spent within `candidate_backing::process_second`",
+					)
+				)?,
+				registry,
+			)?,
+			process_statement: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_candidate_backing_process_statement",
+						"Time spent within `candidate_backing::process_statement`",
+					)
+				)?,
+				registry,
+			)?,
+			get_backed_candidates: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_candidate_backing_get_backed_candidates",
+						"Time spent within `candidate_backing::get_backed_candidates`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}

polkadot/node/core/bitfield-signing/src/lib.rs

@@ -230,6 +230,7 @@ async fn construct_availability_bitfield(
 #[derive(Clone)]
 struct MetricsInner {
 	bitfields_signed_total: prometheus::Counter<prometheus::U64>,
+	run: prometheus::Histogram,
 }
 
 /// Bitfield signing metrics.
@@ -242,6 +243,11 @@ impl Metrics {
 			metrics.bitfields_signed_total.inc();
 		}
 	}
+
+	/// Provide a timer for `prune_povs` which observes on drop.
+	fn time_run(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.run.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -254,6 +260,15 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			run: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_bitfield_signing_run",
+						"Time spent within `bitfield_signing::run`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}
@@ -277,6 +292,7 @@ impl JobTrait for BitfieldSigningJob {
 		_receiver: mpsc::Receiver<ToJob>,
 		mut sender: mpsc::Sender<FromJob>,
 	) -> Pin<Box<dyn Future<Output = Result<(), Self::Error>> + Send>> {
+		let metrics = metrics.clone();
 		async move {
 			let wait_until = Instant::now() + JOB_DELAY;
 
@@ -291,6 +307,10 @@ impl JobTrait for BitfieldSigningJob {
 			// wait a bit before doing anything else
 			Delay::new_at(wait_until).await?;
 
+			// this timer does not appear at the head of the function because we don't want to include
+			// JOB_DELAY each time.
+			let _timer = metrics.time_run();
+
 			let bitfield =
 				match construct_availability_bitfield(relay_parent, validator.index(), &mut sender).await
 			{

polkadot/node/core/candidate-selection/src/lib.rs

@@ -204,6 +204,8 @@ impl CandidateSelectionJob {
 		para_id: ParaId,
 		collator_id: CollatorId,
 	) {
+		let _timer = self.metrics.time_handle_collation();
+
 		if self.seconded_candidate.is_none() {
 			let (candidate_receipt, pov) =
 				match get_collation(
@@ -240,6 +242,8 @@ impl CandidateSelectionJob {
 
 	#[tracing::instrument(level = "trace", skip(self), fields(subsystem = LOG_TARGET))]
 	async fn handle_invalid(&mut self, candidate_receipt: CandidateReceipt) {
+		let _timer = self.metrics.time_handle_invalid();
+
 		let received_from = match &self.seconded_candidate {
 			Some(peer) => peer,
 			None => {
@@ -336,6 +340,8 @@ async fn forward_invalidity_note(
 struct MetricsInner {
 	seconds: prometheus::CounterVec<prometheus::U64>,
 	invalid_selections: prometheus::CounterVec<prometheus::U64>,
+	handle_collation: prometheus::Histogram,
+	handle_invalid: prometheus::Histogram,
 }
 
 /// Candidate selection metrics.
@@ -356,6 +362,16 @@ impl Metrics {
 			metrics.invalid_selections.with_label_values(&[label]).inc();
 		}
 	}
+
+	/// Provide a timer for `handle_collation` which observes on drop.
+	fn time_handle_collation(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.handle_collation.start_timer())
+	}
+
+	/// Provide a timer for `handle_invalid` which observes on drop.
+	fn time_handle_invalid(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.handle_invalid.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -381,6 +397,24 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			handle_collation: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_candidate_selection_handle_collation",
+						"Time spent within `candidate_selection::handle_collation`",
+					)
+				)?,
+				registry,
+			)?,
+			handle_invalid: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_candidate_selection:handle_invalid",
+						"Time spent within `candidate_selection::handle_invalid`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}

polkadot/node/core/candidate-validation/src/lib.rs

@@ -103,12 +103,15 @@ async fn run(
 					pov,
 					response_sender,
 				) => {
+					let _timer = metrics.time_validate_from_chain_state();
+
 					let res = spawn_validate_from_chain_state(
 						&mut ctx,
 						isolation_strategy.clone(),
 						descriptor,
 						pov,
 						spawn.clone(),
+						&metrics,
 					).await;
 
 					match res {
@@ -126,6 +129,8 @@ async fn run(
 					pov,
 					response_sender,
 				) => {
+					let _timer = metrics.time_validate_from_exhaustive();
+
 					let res = spawn_validate_exhaustive(
 						&mut ctx,
 						isolation_strategy.clone(),
@@ -134,6 +139,7 @@ async fn run(
 						descriptor,
 						pov,
 						spawn.clone(),
+						&metrics,
 					).await;
 
 					match res {
@@ -260,13 +266,14 @@ async fn find_assumed_validation_data(
 	Ok(AssumptionCheckOutcome::DoesNotMatch)
 }
 
-#[tracing::instrument(level = "trace", skip(ctx, pov, spawn), fields(subsystem = LOG_TARGET))]
+#[tracing::instrument(level = "trace", skip(ctx, pov, spawn, metrics), fields(subsystem = LOG_TARGET))]
 async fn spawn_validate_from_chain_state(
 	ctx: &mut impl SubsystemContext<Message = CandidateValidationMessage>,
 	isolation_strategy: IsolationStrategy,
 	descriptor: CandidateDescriptor,
 	pov: Arc<PoV>,
 	spawn: impl SpawnNamed + 'static,
+	metrics: &Metrics,
 ) -> SubsystemResult<Result<ValidationResult, ValidationFailed>> {
 	let (validation_data, validation_code) =
 		match find_assumed_validation_data(ctx, &descriptor).await? {
@@ -292,6 +299,7 @@ async fn spawn_validate_from_chain_state(
 		descriptor.clone(),
 		pov,
 		spawn,
+		metrics,
 	)
 	.await;
 
@@ -320,7 +328,7 @@ async fn spawn_validate_from_chain_state(
 	validation_result
 }
 
-#[tracing::instrument(level = "trace", skip(ctx, validation_code, pov, spawn), fields(subsystem = LOG_TARGET))]
+#[tracing::instrument(level = "trace", skip(ctx, validation_code, pov, spawn, metrics), fields(subsystem = LOG_TARGET))]
 async fn spawn_validate_exhaustive(
 	ctx: &mut impl SubsystemContext<Message = CandidateValidationMessage>,
 	isolation_strategy: IsolationStrategy,
@@ -329,8 +337,10 @@ async fn spawn_validate_exhaustive(
 	descriptor: CandidateDescriptor,
 	pov: Arc<PoV>,
 	spawn: impl SpawnNamed + 'static,
+	metrics: &Metrics,
 ) -> SubsystemResult<Result<ValidationResult, ValidationFailed>> {
 	let (tx, rx) = oneshot::channel();
+	let metrics = metrics.clone();
 	let fut = async move {
 		let res = validate_candidate_exhaustive::<RealValidationBackend, _>(
 			isolation_strategy,
@@ -339,6 +349,7 @@ async fn spawn_validate_exhaustive(
 			descriptor,
 			pov,
 			spawn,
+			&metrics,
 		);
 
 		let _ = tx.send(res);
@@ -408,7 +419,7 @@ impl ValidationBackend for RealValidationBackend {
 /// Validates the candidate from exhaustive parameters.
 ///
 /// Sends the result of validation on the channel once complete.
-#[tracing::instrument(level = "trace", skip(backend_arg, validation_code, pov, spawn), fields(subsystem = LOG_TARGET))]
+#[tracing::instrument(level = "trace", skip(backend_arg, validation_code, pov, spawn, metrics), fields(subsystem = LOG_TARGET))]
 fn validate_candidate_exhaustive<B: ValidationBackend, S: SpawnNamed + 'static>(
 	backend_arg: B::Arg,
 	persisted_validation_data: PersistedValidationData,
@@ -416,7 +427,10 @@ fn validate_candidate_exhaustive<B: ValidationBackend, S: SpawnNamed + 'static>(
 	descriptor: CandidateDescriptor,
 	pov: Arc<PoV>,
 	spawn: S,
+	metrics: &Metrics,
 ) -> Result<ValidationResult, ValidationFailed> {
+	let _timer = metrics.time_validate_candidate_exhaustive();
+
 	if let Err(e) = perform_basic_checks(&descriptor, persisted_validation_data.max_pov_size, &*pov) {
 		return Ok(ValidationResult::Invalid(e))
 	}
@@ -460,6 +474,9 @@ fn validate_candidate_exhaustive<B: ValidationBackend, S: SpawnNamed + 'static>(
 #[derive(Clone)]
 struct MetricsInner {
 	validation_requests: prometheus::CounterVec<prometheus::U64>,
+	validate_from_chain_state: prometheus::Histogram,
+	validate_from_exhaustive: prometheus::Histogram,
+	validate_candidate_exhaustive: prometheus::Histogram,
 }
 
 /// Candidate validation metrics.
@@ -482,6 +499,21 @@ impl Metrics {
 			}
 		}
 	}
+
+	/// Provide a timer for `validate_from_chain_state` which observes on drop.
+	fn time_validate_from_chain_state(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.validate_from_chain_state.start_timer())
+	}
+
+	/// Provide a timer for `validate_from_exhaustive` which observes on drop.
+	fn time_validate_from_exhaustive(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.validate_from_exhaustive.start_timer())
+	}
+
+	/// Provide a timer for `validate_candidate_exhaustive` which observes on drop.
+	fn time_validate_candidate_exhaustive(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.validate_candidate_exhaustive.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -497,6 +529,33 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			validate_from_chain_state: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_candidate_validation_validate_from_chain_state",
+						"Time spent within `candidate_validation::validate_from_chain_state`",
+					)
+				)?,
+				registry,
+			)?,
+			validate_from_exhaustive: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_candidate_validation_validate_from_exhaustive",
+						"Time spent within `candidate_validation::validate_from_exhaustive`",
+					)
+				)?,
+				registry,
+			)?,
+			validate_candidate_exhaustive: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_candidate_validation_validate_candidate_exhaustive",
+						"Time spent within `candidate_validation::validate_candidate_exhaustive`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}
@@ -850,6 +909,7 @@ mod tests {
 			descriptor,
 			Arc::new(pov),
 			TaskExecutor::new(),
+			&Default::default(),
 		).unwrap();
 
 		assert_matches!(v, ValidationResult::Valid(outputs, used_validation_data) => {
@@ -885,6 +945,7 @@ mod tests {
 			descriptor,
 			Arc::new(pov),
 			TaskExecutor::new(),
+			&Default::default(),
 		).unwrap();
 
 		assert_matches!(v, ValidationResult::Invalid(InvalidCandidate::BadReturn));
@@ -913,6 +974,7 @@ mod tests {
 			descriptor,
 			Arc::new(pov),
 			TaskExecutor::new(),
+			&Default::default(),
 		);
 
 		assert_matches!(v, Ok(ValidationResult::Invalid(InvalidCandidate::Timeout)));

polkadot/node/core/chain-api/src/lib.rs

@@ -92,11 +92,13 @@ where
 			FromOverseer::Signal(OverseerSignal::BlockFinalized(_)) => {},
 			FromOverseer::Communication { msg } => match msg {
 				ChainApiMessage::BlockNumber(hash, response_channel) => {
+					let _timer = subsystem.metrics.time_block_number();
 					let result = subsystem.client.number(hash).map_err(|e| e.to_string().into());
 					subsystem.metrics.on_request(result.is_ok());
 					let _ = response_channel.send(result);
 				},
 				ChainApiMessage::BlockHeader(hash, response_channel) => {
+					let _timer = subsystem.metrics.time_block_header();
 					let result = subsystem.client
 						.header(BlockId::Hash(hash))
 						.map_err(|e| e.to_string().into());
@@ -104,19 +106,23 @@ where
 					let _ = response_channel.send(result);
 				},
 				ChainApiMessage::FinalizedBlockHash(number, response_channel) => {
+					let _timer = subsystem.metrics.time_finalized_block_hash();
 					// Note: we don't verify it's finalized
 					let result = subsystem.client.hash(number).map_err(|e| e.to_string().into());
 					subsystem.metrics.on_request(result.is_ok());
 					let _ = response_channel.send(result);
 				},
 				ChainApiMessage::FinalizedBlockNumber(response_channel) => {
+					let _timer = subsystem.metrics.time_finalized_block_number();
 					let result = subsystem.client.info().finalized_number;
 					// always succeeds
 					subsystem.metrics.on_request(true);
 					let _ = response_channel.send(Ok(result));
 				},
 				ChainApiMessage::Ancestors { hash, k, response_channel } => {
+					let _timer = subsystem.metrics.time_ancestors();
 					tracing::span!(tracing::Level::TRACE, "ChainApiMessage::Ancestors", subsystem=LOG_TARGET, hash=%hash, k=k);
+
 					let mut hash = hash;
 
 					let next_parent = core::iter::from_fn(|| {
@@ -153,6 +159,11 @@ where
 #[derive(Clone)]
 struct MetricsInner {
 	chain_api_requests: prometheus::CounterVec<prometheus::U64>,
+	block_number: prometheus::Histogram,
+	block_header: prometheus::Histogram,
+	finalized_block_hash: prometheus::Histogram,
+	finalized_block_number: prometheus::Histogram,
+	ancestors: prometheus::Histogram,
 }
 
 /// Chain API metrics.
@@ -169,6 +180,31 @@ impl Metrics {
 			}
 		}
 	}
+
+	/// Provide a timer for `block_number` which observes on drop.
+	fn time_block_number(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.block_number.start_timer())
+	}
+
+	/// Provide a timer for `block_header` which observes on drop.
+	fn time_block_header(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.block_header.start_timer())
+	}
+
+	/// Provide a timer for `finalized_block_hash` which observes on drop.
+	fn time_finalized_block_hash(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.finalized_block_hash.start_timer())
+	}
+
+	/// Provide a timer for `finalized_block_number` which observes on drop.
+	fn time_finalized_block_number(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.finalized_block_number.start_timer())
+	}
+
+	/// Provide a timer for `ancestors` which observes on drop.
+	fn time_ancestors(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.ancestors.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -184,6 +220,51 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			block_number: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_chain_api_block_number",
+						"Time spent within `chain_api::block_number`",
+					)
+				)?,
+				registry,
+			)?,
+			block_header: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_chain_api_block_headers",
+						"Time spent within `chain_api::block_headers`",
+					)
+				)?,
+				registry,
+			)?,
+			finalized_block_hash: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_chain_api_finalized_block_hash",
+						"Time spent within `chain_api::finalized_block_hash`",
+					)
+				)?,
+				registry,
+			)?,
+			finalized_block_number: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_chain_api_finalized_block_number",
+						"Time spent within `chain_api::finalized_block_number`",
+					)
+				)?,
+				registry,
+			)?,
+			ancestors: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_chain_api_ancestors",
+						"Time spent within `chain_api::ancestors`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}

polkadot/node/core/provisioner/src/lib.rs

@@ -197,6 +197,8 @@ impl ProvisioningJob {
 
 			match msg {
 				ToJob::Provisioner(RequestInherentData(_, return_sender)) => {
+					let _timer = self.metrics.time_request_inherent_data();
+
 					if let Err(err) = send_inherent_data(
 						self.relay_parent,
 						&self.signed_bitfields,
@@ -216,6 +218,8 @@ impl ProvisioningJob {
 					self.provisionable_data_channels.push(sender)
 				}
 				ToJob::Provisioner(ProvisionableData(_, data)) => {
+					let _timer = self.metrics.time_provisionable_data();
+
 					let mut bad_indices = Vec::new();
 					for (idx, channel) in self.provisionable_data_channels.iter_mut().enumerate() {
 						match channel.send(data.clone()).await {
@@ -488,6 +492,8 @@ fn bitfields_indicate_availability(
 #[derive(Clone)]
 struct MetricsInner {
 	inherent_data_requests: prometheus::CounterVec<prometheus::U64>,
+	request_inherent_data: prometheus::Histogram,
+	provisionable_data: prometheus::Histogram,
 }
 
 /// Provisioner metrics.
@@ -503,6 +509,16 @@ impl Metrics {
 			}
 		}
 	}
+
+	/// Provide a timer for `request_inherent_data` which observes on drop.
+	fn time_request_inherent_data(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.request_inherent_data.start_timer())
+	}
+
+	/// Provide a timer for `provisionable_data` which observes on drop.
+	fn time_provisionable_data(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.provisionable_data.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -518,6 +534,24 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			request_inherent_data: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_provisioner_request_inherent_data",
+						"Time spent within `provisioner::request_inherent_data`",
+					)
+				)?,
+				registry,
+			)?,
+			provisionable_data: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_provisioner_provisionable_data",
+						"Time spent within `provisioner::provisionable_data`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}

polkadot/node/core/runtime-api/src/lib.rs

@@ -103,6 +103,8 @@ fn make_runtime_api_request<Client>(
 	Client: ProvideRuntimeApi<Block>,
 	Client::Api: ParachainHost<Block>,
 {
+	let _timer = metrics.time_make_runtime_api_request();
+
 	macro_rules! query {
 		($api_name:ident ($($param:expr),*), $sender:expr) => {{
 			let sender = $sender;
@@ -141,6 +143,7 @@ fn make_runtime_api_request<Client>(
 #[derive(Clone)]
 struct MetricsInner {
 	chain_api_requests: prometheus::CounterVec<prometheus::U64>,
+	make_runtime_api_request: prometheus::Histogram,
 }
 
 /// Runtime API metrics.
@@ -157,6 +160,11 @@ impl Metrics {
 			}
 		}
 	}
+
+	/// Provide a timer for `make_runtime_api_request` which observes on drop.
+	fn time_make_runtime_api_request(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.make_runtime_api_request.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -172,6 +180,15 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			make_runtime_api_request: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_runtime_api_make_runtime_api_request",
+						"Time spent within `runtime_api::make_runtime_api_request`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}

polkadot/node/network/availability-distribution/src/lib.rs

@@ -385,6 +385,8 @@ async fn handle_our_view_change<Context>(
 where
 	Context: SubsystemContext<Message = AvailabilityDistributionMessage>,
 {
+	let _timer = metrics.time_handle_our_view_change();
+
 	let old_view = std::mem::replace(&mut (state.view), view);
 
 	// needed due to borrow rules
@@ -651,6 +653,8 @@ async fn process_incoming_peer_message<Context>(
 where
 	Context: SubsystemContext<Message = AvailabilityDistributionMessage>,
 {
+	let _timer = metrics.time_process_incoming_peer_message();
+
 	// obtain the set of candidates we are interested in based on our current view
 	let live_candidates = state.cached_live_candidates_unioned(state.view.0.iter());
 
@@ -1180,6 +1184,8 @@ where
 #[derive(Clone)]
 struct MetricsInner {
 	gossipped_availability_chunks: prometheus::Counter<prometheus::U64>,
+	handle_our_view_change: prometheus::Histogram,
+	process_incoming_peer_message: prometheus::Histogram,
 }
 
 /// Availability Distribution metrics.
@@ -1192,6 +1198,16 @@ impl Metrics {
 			metrics.gossipped_availability_chunks.inc();
 		}
 	}
+
+	/// Provide a timer for `handle_our_view_change` which observes on drop.
+	fn time_handle_our_view_change(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.handle_our_view_change.start_timer())
+	}
+
+	/// Provide a timer for `process_incoming_peer_message` which observes on drop.
+	fn time_process_incoming_peer_message(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.process_incoming_peer_message.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -1206,6 +1222,24 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			handle_our_view_change: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_availability_distribution_handle_our_view_change",
+						"Time spent within `availability_distribution::handle_our_view_change`",
+					)
+				)?,
+				registry,
+			)?,
+			process_incoming_peer_message: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_availability_distribution_process_incoming_peer_message",
+						"Time spent within `availability_distribution::process_incoming_peer_message`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}

polkadot/node/network/bitfield-distribution/src/lib.rs

@@ -183,6 +183,8 @@ impl BitfieldDistribution {
 					}
 				}
 				FromOverseer::Signal(OverseerSignal::ActiveLeaves(ActiveLeavesUpdate { activated, deactivated })) => {
+					let _timer = self.metrics.time_active_leaves_update();
+
 					for relay_parent in activated {
 						tracing::trace!(target: LOG_TARGET, relay_parent = %relay_parent, "activated");
 						// query basic system parameters once
@@ -257,6 +259,8 @@ async fn handle_bitfield_distribution<Context>(
 where
 	Context: SubsystemContext<Message = BitfieldDistributionMessage>,
 {
+	let _timer = metrics.time_handle_bitfield_distribution();
+
 	// Ignore anything the overseer did not tell this subsystem to work on
 	let mut job_data = state.per_relay_parent.get_mut(&relay_parent);
 	let job_data: &mut _ = if let Some(ref mut job_data) = job_data {
@@ -461,6 +465,8 @@ async fn handle_network_msg<Context>(
 where
 	Context: SubsystemContext<Message = BitfieldDistributionMessage>,
 {
+	let _timer = metrics.time_handle_network_msg();
+
 	match bridge_message {
 		NetworkBridgeEvent::PeerConnected(peerid, _role) => {
 			// insert if none already present
@@ -659,6 +665,9 @@ where
 struct MetricsInner {
 	gossipped_own_availability_bitfields: prometheus::Counter<prometheus::U64>,
 	received_availability_bitfields: prometheus::Counter<prometheus::U64>,
+	active_leaves_update: prometheus::Histogram,
+	handle_bitfield_distribution: prometheus::Histogram,
+	handle_network_msg: prometheus::Histogram,
 }
 
 /// Bitfield Distribution metrics.
@@ -677,6 +686,21 @@ impl Metrics {
 			metrics.received_availability_bitfields.inc();
 		}
 	}
+
+	/// Provide a timer for `active_leaves_update` which observes on drop.
+	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 `handle_bitfield_distribution` which observes on drop.
+	fn time_handle_bitfield_distribution(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.handle_bitfield_distribution.start_timer())
+	}
+
+	/// Provide a timer for `handle_network_msg` which observes on drop.
+	fn time_handle_network_msg(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.handle_network_msg.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -696,6 +720,33 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			active_leaves_update: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_bitfield_distribution_active_leaves_update",
+						"Time spent within `bitfield_distribution::active_leaves_update`",
+					)
+				)?,
+				registry,
+			)?,
+			handle_bitfield_distribution: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_bitfield_distribution_handle_bitfield_distribution",
+						"Time spent within `bitfield_distribution::handle_bitfield_distribution`",
+					)
+				)?,
+				registry,
+			)?,
+			handle_network_msg: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_bitfield_distribution_handle_network_msg",
+						"Time spent within `bitfield_distribution::handle_network_msg`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}

polkadot/node/network/collator-protocol/src/collator_side.rs

@@ -57,12 +57,24 @@ impl Metrics {
 			metrics.collations_sent.inc();
 		}
 	}
+
+	/// Provide a timer for handling `ConnectionRequest` which observes on drop.
+	fn time_handle_connection_request(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.handle_connection_request.start_timer())
+	}
+
+	/// Provide a timer for `process_msg` which observes on drop.
+	fn time_process_msg(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.process_msg.start_timer())
+	}
 }
 
 #[derive(Clone)]
 struct MetricsInner {
 	advertisements_made: prometheus::Counter<prometheus::U64>,
 	collations_sent: prometheus::Counter<prometheus::U64>,
+	handle_connection_request: prometheus::Histogram,
+	process_msg: prometheus::Histogram,
 }
 
 impl metrics::Metrics for Metrics {
@@ -84,6 +96,24 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			handle_connection_request: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_collator_protocol_collator_handle_connection_request",
+						"Time spent within `collator_protocol_collator::handle_connection_request`",
+					)
+				)?,
+				registry,
+			)?,
+			process_msg: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_collator_protocol_collator_process_msg",
+						"Time spent within `collator_protocol_collator::process_msg`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 
 		Ok(Metrics(Some(metrics)))
@@ -377,6 +407,8 @@ where
 {
 	use CollatorProtocolMessage::*;
 
+	let _timer = state.metrics.time_process_msg();
+
 	match msg {
 		CollateOn(id) => {
 			state.collating_on = Some(id);
@@ -662,6 +694,8 @@ where
 
 	loop {
 		if let Some(mut request) = state.last_connection_request.take() {
+			let _timer = state.metrics.time_handle_connection_request();
+
 			while let Poll::Ready(Some((validator_id, peer_id))) = futures::poll!(request.next()) {
 				if let Err(err) = handle_validator_connected(&mut ctx, &mut state, peer_id, validator_id).await {
 					tracing::warn!(

polkadot/node/network/collator-protocol/src/validator_side.rs

@@ -63,11 +63,23 @@ impl Metrics {
 			}
 		}
 	}
+
+	/// Provide a timer for `process_msg` which observes on drop.
+	fn time_process_msg(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.process_msg.start_timer())
+	}
+
+	/// Provide a timer for `handle_collation_request_result` which observes on drop.
+	fn time_handle_collation_request_result(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.handle_collation_request_result.start_timer())
+	}
 }
 
 #[derive(Clone)]
 struct MetricsInner {
 	collation_requests: prometheus::CounterVec<prometheus::U64>,
+	process_msg: prometheus::Histogram,
+	handle_collation_request_result: prometheus::Histogram,
 }
 
 impl metrics::Metrics for Metrics {
@@ -84,7 +96,25 @@ impl metrics::Metrics for Metrics {
 					&["success"],
 				)?,
 				registry,
-			)?
+			)?,
+			process_msg: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_collator_protocol_validator_process_msg",
+						"Time spent within `collator_protocol_validator::process_msg`",
+					)
+				)?,
+				registry,
+			)?,
+			handle_collation_request_result: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_collator_protocol_validator_handle_collation_request_result",
+						"Time spent within `collator_protocol_validator::handle_collation_request_result`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 
 		Ok(Metrics(Some(metrics)))
@@ -628,6 +658,8 @@ where
 {
 	use CollatorProtocolMessage::*;
 
+	let _timer = state.metrics.time_process_msg();
+
 	match msg {
 		CollateOn(id) => {
 			tracing::warn!(
@@ -703,6 +735,8 @@ where
 		}
 
 		while let Poll::Ready(Some(request)) = futures::poll!(state.requests_in_progress.next()) {
+			let _timer = state.metrics.time_handle_collation_request_result();
+
 			// Request has timed out, we need to penalize the collator and re-send the request
 			// if the chain has not moved on yet.
 			match request {

polkadot/node/network/pov-distribution/src/lib.rs

@@ -125,6 +125,8 @@ async fn handle_signal(
 	match signal {
 		OverseerSignal::Conclude => Ok(true),
 		OverseerSignal::ActiveLeaves(ActiveLeavesUpdate { activated, deactivated }) => {
+			let _timer = state.metrics.time_handle_signal();
+
 			for relay_parent in activated {
 				let (vals_tx, vals_rx) = oneshot::channel();
 				ctx.send_message(AllMessages::RuntimeApi(RuntimeApiMessage::Request(
@@ -267,6 +269,8 @@ async fn handle_fetch(
 	descriptor: CandidateDescriptor,
 	response_sender: oneshot::Sender<Arc<PoV>>,
 ) -> SubsystemResult<()> {
+	let _timer = state.metrics.time_handle_fetch();
+
 	let relay_parent_state = match state.relay_parent_state.get_mut(&relay_parent) {
 		Some(s) => s,
 		None => return Ok(()),
@@ -316,6 +320,8 @@ async fn handle_distribute(
 	descriptor: CandidateDescriptor,
 	pov: Arc<PoV>,
 ) -> SubsystemResult<()> {
+	let _timer = state.metrics.time_handle_distribute();
+
 	let relay_parent_state = match state.relay_parent_state.get_mut(&relay_parent) {
 		None => return Ok(()),
 		Some(s) => s,
@@ -483,6 +489,8 @@ async fn handle_network_update(
 	ctx: &mut impl SubsystemContext<Message = PoVDistributionMessage>,
 	update: NetworkBridgeEvent<protocol_v1::PoVDistributionMessage>,
 ) -> SubsystemResult<()> {
+	let _timer = state.metrics.time_handle_network_update();
+
 	match update {
 		NetworkBridgeEvent::PeerConnected(peer, _observed_role) => {
 			state.peer_state.insert(peer, PeerState { awaited: HashMap::new() });
@@ -600,6 +608,10 @@ impl PoVDistribution {
 #[derive(Clone)]
 struct MetricsInner {
 	povs_distributed: prometheus::Counter<prometheus::U64>,
+	handle_signal: prometheus::Histogram,
+	handle_fetch: prometheus::Histogram,
+	handle_distribute: prometheus::Histogram,
+	handle_network_update: prometheus::Histogram,
 }
 
 /// Availability Distribution metrics.
@@ -612,6 +624,26 @@ impl Metrics {
 			metrics.povs_distributed.inc();
 		}
 	}
+
+	/// Provide a timer for `handle_signal` which observes on drop.
+	fn time_handle_signal(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.handle_signal.start_timer())
+	}
+
+	/// Provide a timer for `handle_fetch` which observes on drop.
+	fn time_handle_fetch(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.handle_fetch.start_timer())
+	}
+
+	/// Provide a timer for `handle_distribute` which observes on drop.
+	fn time_handle_distribute(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.handle_distribute.start_timer())
+	}
+
+	/// Provide a timer for `handle_network_update` which observes on drop.
+	fn time_handle_network_update(&self) -> Option<metrics::prometheus::prometheus::HistogramTimer> {
+		self.0.as_ref().map(|metrics| metrics.handle_network_update.start_timer())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -624,6 +656,42 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			handle_signal: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_pov_distribution_handle_signal",
+						"Time spent within `pov_distribution::handle_signal`",
+					)
+				)?,
+				registry,
+			)?,
+			handle_fetch: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_pov_distribution_handle_fetch",
+						"Time spent within `pov_distribution::handle_fetch`",
+					)
+				)?,
+				registry,
+			)?,
+			handle_distribute: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_pov_distribution_handle_distribute",
+						"Time spent within `pov_distribution::handle_distribute`",
+					)
+				)?,
+				registry,
+			)?,
+			handle_network_update: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_pov_distribution_handle_network_update",
+						"Time spent within `pov_distribution::handle_network_update`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}

polkadot/node/network/statement-distribution/src/lib.rs

@@ -895,6 +895,8 @@ impl StatementDistribution {
 			let message = ctx.recv().await?;
 			match message {
 				FromOverseer::Signal(OverseerSignal::ActiveLeaves(ActiveLeavesUpdate { activated, .. })) => {
+					let _timer = metrics.time_active_leaves_update();
+
 					for relay_parent in activated {
 						let (validators, session_index) = {
 							let (val_tx, val_rx) = oneshot::channel();
@@ -944,6 +946,8 @@ impl StatementDistribution {
 				FromOverseer::Signal(OverseerSignal::Conclude) => break,
 				FromOverseer::Communication { msg } => match msg {
 					StatementDistributionMessage::Share(relay_parent, statement) => {
+						let _timer = metrics.time_share();
+
 						inform_statement_listeners(
 							&statement,
 							&mut statement_listeners,
@@ -957,7 +961,9 @@ impl StatementDistribution {
 							&metrics,
 						).await?;
 					}
-					StatementDistributionMessage::NetworkBridgeUpdateV1(event) =>
+					StatementDistributionMessage::NetworkBridgeUpdateV1(event) => {
+						let _timer = metrics.time_network_bridge_update_v1();
+
 						handle_network_update(
 							&mut peers,
 							&mut active_heads,
@@ -965,7 +971,8 @@ impl StatementDistribution {
 							&mut our_view,
 							event,
 							&metrics,
-						).await?,
+						).await?
+					}
 					StatementDistributionMessage::RegisterStatementListener(tx) => {
 						statement_listeners.push(tx);
 					}
@@ -979,6 +986,9 @@ impl StatementDistribution {
 #[derive(Clone)]
 struct MetricsInner {
 	statements_distributed: prometheus::Counter<prometheus::U64>,
+	active_leaves_update: prometheus::Histogram,
+	share: prometheus::Histogram,
+	network_bridge_update_v1: prometheus::Histogram,
 }
 
 /// Statement Distribution metrics.
@@ -991,6 +1001,21 @@ impl Metrics {
 			metrics.statements_distributed.inc();
 		}
 	}
+
+	/// Provide a timer for `active_leaves_update` which observes on drop.
+	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.
+	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.
+	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())
+	}
 }
 
 impl metrics::Metrics for Metrics {
@@ -1003,6 +1028,33 @@ impl metrics::Metrics for Metrics {
 				)?,
 				registry,
 			)?,
+			active_leaves_update: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"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(
+						"parachain_statement_distribution_share",
+						"Time spent within `statement_distribution::share`",
+					)
+				)?,
+				registry,
+			)?,
+			network_bridge_update_v1: prometheus::register(
+				prometheus::Histogram::with_opts(
+					prometheus::HistogramOpts::new(
+						"parachain_statement_distribution_network_bridge_update_v1",
+						"Time spent within `statement_distribution::network_bridge_update_v1`",
+					)
+				)?,
+				registry,
+			)?,
 		};
 		Ok(Metrics(Some(metrics)))
 	}

polkadot/node/subsystem-util/src/lib.rs

@@ -442,9 +442,10 @@ impl<ToJob: ToJobTrait> JobHandle<ToJob> {
 
 /// This module reexports Prometheus types and defines the [`Metrics`] trait.
 pub mod metrics {
-	/// Reexport Prometheus types.
+	/// Reexport Substrate Prometheus types.
 	pub use substrate_prometheus_endpoint as prometheus;
 
+
 	/// Subsystem- or job-specific Prometheus metrics.
 	///
 	/// Usually implemented as a wrapper for `Option<ActualMetrics>`