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
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
// This file is part of Substrate.

// Copyright (C) Parity Technologies (UK) Ltd.
// SPDX-License-Identifier: GPL-3.0-or-later WITH Classpath-exception-2.0

// This program 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.

// This program 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 this program. If not, see <https://www.gnu.org/licenses/>.

use crate::{
	error::{Error, Result},
	interval::ExpIncInterval,
	ServicetoWorkerMsg, WorkerConfig,
};

use std::{
	collections::{HashMap, HashSet},
	marker::PhantomData,
	sync::Arc,
	time::{Duration, Instant, SystemTime, UNIX_EPOCH},
};

use futures::{channel::mpsc, future, stream::Fuse, FutureExt, Stream, StreamExt};

use addr_cache::AddrCache;
use codec::{Decode, Encode};
use ip_network::IpNetwork;
use libp2p::kad::{PeerRecord, Record};
use linked_hash_set::LinkedHashSet;

use log::{debug, error, trace};
use prometheus_endpoint::{register, Counter, CounterVec, Gauge, Opts, U64};
use prost::Message;
use rand::{seq::SliceRandom, thread_rng};

use sc_network::{
	config::DEFAULT_KADEMLIA_REPLICATION_FACTOR, event::DhtEvent, multiaddr, KademliaKey,
	Multiaddr, NetworkDHTProvider, NetworkSigner, NetworkStateInfo,
};
use sc_network_types::{multihash::Code, PeerId};
use schema::PeerSignature;
use sp_api::{ApiError, ProvideRuntimeApi};
use sp_authority_discovery::{
	AuthorityDiscoveryApi, AuthorityId, AuthorityPair, AuthoritySignature,
};
use sp_blockchain::HeaderBackend;
use sp_core::crypto::{key_types, ByteArray, Pair};
use sp_keystore::{Keystore, KeystorePtr};
use sp_runtime::traits::Block as BlockT;

mod addr_cache;
/// Dht payload schemas generated from Protobuf definitions via Prost crate in build.rs.
mod schema {
	#[cfg(test)]
	mod tests;

	include!(concat!(env!("OUT_DIR"), "/authority_discovery_v3.rs"));
}
#[cfg(test)]
pub mod tests;

const LOG_TARGET: &str = "sub-authority-discovery";

/// Maximum number of addresses cached per authority. Additional addresses are discarded.
const MAX_ADDRESSES_PER_AUTHORITY: usize = 10;

/// Maximum number of in-flight DHT lookups at any given point in time.
const MAX_IN_FLIGHT_LOOKUPS: usize = 8;

/// Role an authority discovery [`Worker`] can run as.
pub enum Role {
	/// Publish own addresses and discover addresses of others.
	PublishAndDiscover(KeystorePtr),
	/// Discover addresses of others.
	Discover,
}

/// An authority discovery [`Worker`] can publish the local node's addresses as well as discover
/// those of other nodes via a Kademlia DHT.
///
/// When constructed with [`Role::PublishAndDiscover`] a [`Worker`] will
///
///    1. Retrieve its external addresses (including peer id).
///
///    2. Get the list of keys owned by the local node participating in the current authority set.
///
///    3. Sign the addresses with the keys.
///
///    4. Put addresses and signature as a record with the authority id as a key on a Kademlia DHT.
///
/// When constructed with either [`Role::PublishAndDiscover`] or [`Role::Discover`] a [`Worker`]
/// will
///
///    1. Retrieve the current and next set of authorities.
///
///    2. Start DHT queries for the ids of the authorities.
///
///    3. Validate the signatures of the retrieved key value pairs.
///
///    4. Add the retrieved external addresses as priority nodes to the
///    network peerset.
///
///    5. Allow querying of the collected addresses via the [`crate::Service`].
pub struct Worker<Client, Block: BlockT, DhtEventStream> {
	/// Channel receiver for messages send by a [`crate::Service`].
	from_service: Fuse<mpsc::Receiver<ServicetoWorkerMsg>>,

	client: Arc<Client>,

	network: Arc<dyn NetworkProvider>,

	/// Channel we receive Dht events on.
	dht_event_rx: DhtEventStream,

	/// Interval to be proactive, publishing own addresses.
	publish_interval: ExpIncInterval,

	/// Pro-actively publish our own addresses at this interval, if the keys in the keystore
	/// have changed.
	publish_if_changed_interval: ExpIncInterval,

	/// List of keys onto which addresses have been published at the latest publication.
	/// Used to check whether they have changed.
	latest_published_keys: HashSet<AuthorityId>,
	/// List of the kademlia keys that have been published at the latest publication.
	/// Used to associate DHT events with our published records.
	latest_published_kad_keys: HashSet<KademliaKey>,

	/// Same value as in the configuration.
	publish_non_global_ips: bool,

	/// Public addresses set by the node operator to always publish first in the authority
	/// discovery DHT record.
	public_addresses: LinkedHashSet<Multiaddr>,

	/// Same value as in the configuration.
	strict_record_validation: bool,

	/// Interval at which to request addresses of authorities, refilling the pending lookups queue.
	query_interval: ExpIncInterval,

	/// Queue of throttled lookups pending to be passed to the network.
	pending_lookups: Vec<AuthorityId>,

	/// The list of all known authorities.
	known_authorities: HashMap<KademliaKey, AuthorityId>,

	/// The last time we requested the list of authorities.
	authorities_queried_at: Option<Block::Hash>,

	/// Set of in-flight lookups.
	in_flight_lookups: HashMap<KademliaKey, AuthorityId>,

	/// Set of lookups we can still receive records.
	/// These are the entries in the `in_flight_lookups` for which
	/// we got at least one successfull result.
	known_lookups: HashMap<KademliaKey, AuthorityId>,

	/// Last known record by key, here we always keep the record with
	/// the highest creation time and we don't accept records older than
	/// that.
	last_known_records: HashMap<KademliaKey, RecordInfo>,

	addr_cache: addr_cache::AddrCache,

	metrics: Option<Metrics>,

	role: Role,

	phantom: PhantomData<Block>,
}

#[derive(Debug, Clone)]
struct RecordInfo {
	/// Time since UNIX_EPOCH in nanoseconds.
	creation_time: u128,
	/// Peers that we know have this record, bounded to no more than
	/// DEFAULT_KADEMLIA_REPLICATION_FACTOR(20).
	peers_with_record: HashSet<PeerId>,
	/// The record itself.
	record: Record,
}

/// Wrapper for [`AuthorityDiscoveryApi`](sp_authority_discovery::AuthorityDiscoveryApi). Can be
/// be implemented by any struct without dependency on the runtime.
#[async_trait::async_trait]
pub trait AuthorityDiscovery<Block: BlockT> {
	/// Retrieve authority identifiers of the current and next authority set.
	async fn authorities(&self, at: Block::Hash)
		-> std::result::Result<Vec<AuthorityId>, ApiError>;

	/// Retrieve best block hash
	async fn best_hash(&self) -> std::result::Result<Block::Hash, Error>;
}

#[async_trait::async_trait]
impl<Block, T> AuthorityDiscovery<Block> for T
where
	T: ProvideRuntimeApi<Block> + HeaderBackend<Block> + Send + Sync,
	T::Api: AuthorityDiscoveryApi<Block>,
	Block: BlockT,
{
	async fn authorities(
		&self,
		at: Block::Hash,
	) -> std::result::Result<Vec<AuthorityId>, ApiError> {
		self.runtime_api().authorities(at)
	}

	async fn best_hash(&self) -> std::result::Result<Block::Hash, Error> {
		Ok(self.info().best_hash)
	}
}

impl<Client, Block, DhtEventStream> Worker<Client, Block, DhtEventStream>
where
	Block: BlockT + Unpin + 'static,
	Client: AuthorityDiscovery<Block> + 'static,
	DhtEventStream: Stream<Item = DhtEvent> + Unpin,
{
	/// Construct a [`Worker`].
	pub(crate) fn new(
		from_service: mpsc::Receiver<ServicetoWorkerMsg>,
		client: Arc<Client>,
		network: Arc<dyn NetworkProvider>,
		dht_event_rx: DhtEventStream,
		role: Role,
		prometheus_registry: Option<prometheus_endpoint::Registry>,
		config: WorkerConfig,
	) -> Self {
		// When a node starts up publishing and querying might fail due to various reasons, for
		// example due to being not yet fully bootstrapped on the DHT. Thus one should retry rather
		// sooner than later. On the other hand, a long running node is likely well connected and
		// thus timely retries are not needed. For this reasoning use an exponentially increasing
		// interval for `publish_interval`, `query_interval` and `priority_group_set_interval`
		// instead of a constant interval.
		let publish_interval =
			ExpIncInterval::new(Duration::from_secs(2), config.max_publish_interval);
		let query_interval = ExpIncInterval::new(Duration::from_secs(2), config.max_query_interval);

		// An `ExpIncInterval` is overkill here because the interval is constant, but consistency
		// is more simple.
		let publish_if_changed_interval =
			ExpIncInterval::new(config.keystore_refresh_interval, config.keystore_refresh_interval);

		let addr_cache = AddrCache::new();

		let metrics = match prometheus_registry {
			Some(registry) => match Metrics::register(&registry) {
				Ok(metrics) => Some(metrics),
				Err(e) => {
					error!(target: LOG_TARGET, "Failed to register metrics: {}", e);
					None
				},
			},
			None => None,
		};

		let public_addresses = {
			let local_peer_id = network.local_peer_id();

			config
				.public_addresses
				.into_iter()
				.map(|mut address| {
					if let Some(multiaddr::Protocol::P2p(peer_id)) = address.iter().last() {
						if peer_id != *local_peer_id.as_ref() {
							error!(
								target: LOG_TARGET,
								"Discarding invalid local peer ID in public address {address}.",
							);
						}
						// Always discard `/p2p/...` protocol for proper address comparison (local
						// peer id will be added before publishing).
						address.pop();
					}
					address
				})
				.collect()
		};

		Worker {
			from_service: from_service.fuse(),
			client,
			network,
			dht_event_rx,
			publish_interval,
			known_authorities: Default::default(),
			authorities_queried_at: None,
			publish_if_changed_interval,
			latest_published_keys: HashSet::new(),
			latest_published_kad_keys: HashSet::new(),
			publish_non_global_ips: config.publish_non_global_ips,
			public_addresses,
			strict_record_validation: config.strict_record_validation,
			query_interval,
			pending_lookups: Vec::new(),
			in_flight_lookups: HashMap::new(),
			known_lookups: HashMap::new(),
			addr_cache,
			role,
			metrics,
			phantom: PhantomData,
			last_known_records: HashMap::new(),
		}
	}

	/// Start the worker
	pub async fn run(mut self) {
		loop {
			self.start_new_lookups();

			futures::select! {
				// Process incoming events.
				event = self.dht_event_rx.next().fuse() => {
					if let Some(event) = event {
						self.handle_dht_event(event).await;
					} else {
						// This point is reached if the network has shut down, at which point there is not
						// much else to do than to shut down the authority discovery as well.
						return;
					}
				},
				// Handle messages from [`Service`]. Ignore if sender side is closed.
				msg = self.from_service.select_next_some() => {
					self.process_message_from_service(msg);
				},
				// Publish own addresses.
				only_if_changed = future::select(
					self.publish_interval.next().map(|_| false),
					self.publish_if_changed_interval.next().map(|_| true)
				).map(|e| e.factor_first().0).fuse() => {
					if let Err(e) = self.publish_ext_addresses(only_if_changed).await {
						error!(
							target: LOG_TARGET,
							"Failed to publish external addresses: {}", e,
						);
					}
				},
				// Request addresses of authorities.
				_ = self.query_interval.next().fuse() => {
					if let Err(e) = self.refill_pending_lookups_queue().await {
						error!(
							target: LOG_TARGET,
							"Failed to request addresses of authorities: {}", e,
						);
					}
				},
			}
		}
	}

	fn process_message_from_service(&self, msg: ServicetoWorkerMsg) {
		match msg {
			ServicetoWorkerMsg::GetAddressesByAuthorityId(authority, sender) => {
				let _ = sender.send(
					self.addr_cache.get_addresses_by_authority_id(&authority).map(Clone::clone),
				);
			},
			ServicetoWorkerMsg::GetAuthorityIdsByPeerId(peer_id, sender) => {
				let _ = sender
					.send(self.addr_cache.get_authority_ids_by_peer_id(&peer_id).map(Clone::clone));
			},
		}
	}

	fn addresses_to_publish(&self) -> impl Iterator<Item = Multiaddr> {
		let local_peer_id = self.network.local_peer_id();
		let publish_non_global_ips = self.publish_non_global_ips;
		let addresses = self
			.public_addresses
			.clone()
			.into_iter()
			.chain(self.network.external_addresses().into_iter().filter_map(|mut address| {
				// Make sure the reported external address does not contain `/p2p/...` protocol.
				if let Some(multiaddr::Protocol::P2p(peer_id)) = address.iter().last() {
					if peer_id != *local_peer_id.as_ref() {
						error!(
							target: LOG_TARGET,
							"Network returned external address '{address}' with peer id \
							 not matching the local peer id '{local_peer_id}'.",
						);
						debug_assert!(false);
					}
					address.pop();
				}

				if self.public_addresses.contains(&address) {
					// Already added above.
					None
				} else {
					Some(address)
				}
			}))
			.filter(move |address| {
				if publish_non_global_ips {
					return true
				}

				address.iter().all(|protocol| match protocol {
					// The `ip_network` library is used because its `is_global()` method is stable,
					// while `is_global()` in the standard library currently isn't.
					multiaddr::Protocol::Ip4(ip) if !IpNetwork::from(ip).is_global() => false,
					multiaddr::Protocol::Ip6(ip) if !IpNetwork::from(ip).is_global() => false,
					_ => true,
				})
			})
			.collect::<Vec<_>>();

		if !addresses.is_empty() {
			debug!(
				target: LOG_TARGET,
				"Publishing authority DHT record peer_id='{local_peer_id}' with addresses='{addresses:?}'",
			);
		}

		// The address must include the local peer id.
		addresses
			.into_iter()
			.map(move |a| a.with(multiaddr::Protocol::P2p(*local_peer_id.as_ref())))
	}

	/// Publish own public addresses.
	///
	/// If `only_if_changed` is true, the function has no effect if the list of keys to publish
	/// is equal to `self.latest_published_keys`.
	async fn publish_ext_addresses(&mut self, only_if_changed: bool) -> Result<()> {
		let key_store = match &self.role {
			Role::PublishAndDiscover(key_store) => key_store,
			Role::Discover => return Ok(()),
		};

		let addresses = serialize_addresses(self.addresses_to_publish());
		if addresses.is_empty() {
			trace!(
				target: LOG_TARGET,
				"No addresses to publish. Skipping publication."
			);

			self.publish_interval.set_to_start();
			return Ok(())
		}

		let keys =
			Worker::<Client, Block, DhtEventStream>::get_own_public_keys_within_authority_set(
				key_store.clone(),
				self.client.as_ref(),
			)
			.await?
			.into_iter()
			.collect::<HashSet<_>>();

		if only_if_changed {
			// If the authority keys did not change and the `publish_if_changed_interval` was
			// triggered then do nothing.
			if keys == self.latest_published_keys {
				return Ok(())
			}

			// We have detected a change in the authority keys, reset the timers to
			// publish and gather data faster.
			self.publish_interval.set_to_start();
			self.query_interval.set_to_start();
		}

		if let Some(metrics) = &self.metrics {
			metrics.publish.inc();
			metrics
				.amount_addresses_last_published
				.set(addresses.len().try_into().unwrap_or(std::u64::MAX));
		}

		let serialized_record = serialize_authority_record(addresses, Some(build_creation_time()))?;
		let peer_signature = sign_record_with_peer_id(&serialized_record, &self.network)?;

		let keys_vec = keys.iter().cloned().collect::<Vec<_>>();

		let kv_pairs = sign_record_with_authority_ids(
			serialized_record,
			Some(peer_signature),
			key_store.as_ref(),
			keys_vec,
		)?;

		self.latest_published_kad_keys = kv_pairs.iter().map(|(k, _)| k.clone()).collect();

		for (key, value) in kv_pairs.into_iter() {
			self.network.put_value(key, value);
		}

		self.latest_published_keys = keys;

		Ok(())
	}

	async fn refill_pending_lookups_queue(&mut self) -> Result<()> {
		let best_hash = self.client.best_hash().await?;

		let local_keys = match &self.role {
			Role::PublishAndDiscover(key_store) => key_store
				.sr25519_public_keys(key_types::AUTHORITY_DISCOVERY)
				.into_iter()
				.collect::<HashSet<_>>(),
			Role::Discover => HashSet::new(),
		};

		let mut authorities = self
			.client
			.authorities(best_hash)
			.await
			.map_err(|e| Error::CallingRuntime(e.into()))?
			.into_iter()
			.filter(|id| !local_keys.contains(id.as_ref()))
			.collect::<Vec<_>>();

		self.known_authorities = authorities
			.clone()
			.into_iter()
			.map(|authority| (hash_authority_id(authority.as_ref()), authority))
			.collect::<HashMap<_, _>>();
		self.authorities_queried_at = Some(best_hash);

		self.addr_cache.retain_ids(&authorities);
		let now = Instant::now();
		self.last_known_records.retain(|k, value| {
			self.known_authorities.contains_key(k) && !value.record.is_expired(now)
		});

		authorities.shuffle(&mut thread_rng());
		self.pending_lookups = authorities;
		// Ignore all still in-flight lookups. Those that are still in-flight are likely stalled as
		// query interval ticks are far enough apart for all lookups to succeed.
		self.in_flight_lookups.clear();
		self.known_lookups.clear();

		if let Some(metrics) = &self.metrics {
			metrics
				.requests_pending
				.set(self.pending_lookups.len().try_into().unwrap_or(std::u64::MAX));
		}

		Ok(())
	}

	fn start_new_lookups(&mut self) {
		while self.in_flight_lookups.len() < MAX_IN_FLIGHT_LOOKUPS {
			let authority_id = match self.pending_lookups.pop() {
				Some(authority) => authority,
				None => return,
			};
			let hash = hash_authority_id(authority_id.as_ref());
			self.network.get_value(&hash);
			self.in_flight_lookups.insert(hash, authority_id);

			if let Some(metrics) = &self.metrics {
				metrics.requests.inc();
				metrics
					.requests_pending
					.set(self.pending_lookups.len().try_into().unwrap_or(std::u64::MAX));
			}
		}
	}

	/// Handle incoming Dht events.
	async fn handle_dht_event(&mut self, event: DhtEvent) {
		match event {
			DhtEvent::ValueFound(v) => {
				if let Some(metrics) = &self.metrics {
					metrics.dht_event_received.with_label_values(&["value_found"]).inc();
				}

				debug!(target: LOG_TARGET, "Value for hash '{:?}' found on Dht.", v.record.key);

				if let Err(e) = self.handle_dht_value_found_event(v) {
					if let Some(metrics) = &self.metrics {
						metrics.handle_value_found_event_failure.inc();
					}
					debug!(target: LOG_TARGET, "Failed to handle Dht value found event: {}", e);
				}
			},
			DhtEvent::ValueNotFound(hash) => {
				if let Some(metrics) = &self.metrics {
					metrics.dht_event_received.with_label_values(&["value_not_found"]).inc();
				}

				if self.in_flight_lookups.remove(&hash).is_some() {
					debug!(target: LOG_TARGET, "Value for hash '{:?}' not found on Dht.", hash)
				} else {
					debug!(
						target: LOG_TARGET,
						"Received 'ValueNotFound' for unexpected hash '{:?}'.", hash
					)
				}
			},
			DhtEvent::ValuePut(hash) => {
				if !self.latest_published_kad_keys.contains(&hash) {
					return;
				}

				// Fast forward the exponentially increasing interval to the configured maximum. In
				// case this was the first successful address publishing there is no need for a
				// timely retry.
				self.publish_interval.set_to_max();

				if let Some(metrics) = &self.metrics {
					metrics.dht_event_received.with_label_values(&["value_put"]).inc();
				}

				debug!(target: LOG_TARGET, "Successfully put hash '{:?}' on Dht.", hash)
			},
			DhtEvent::ValuePutFailed(hash) => {
				if !self.latest_published_kad_keys.contains(&hash) {
					// Not a value we have published or received multiple times.
					return;
				}

				if let Some(metrics) = &self.metrics {
					metrics.dht_event_received.with_label_values(&["value_put_failed"]).inc();
				}

				debug!(target: LOG_TARGET, "Failed to put hash '{:?}' on Dht.", hash)
			},
			DhtEvent::PutRecordRequest(record_key, record_value, publisher, expires) => {
				if let Err(e) = self
					.handle_put_record_requested(record_key, record_value, publisher, expires)
					.await
				{
					debug!(target: LOG_TARGET, "Failed to handle put record request: {}", e)
				}

				if let Some(metrics) = &self.metrics {
					metrics.dht_event_received.with_label_values(&["put_record_req"]).inc();
				}
			},
		}
	}

	async fn handle_put_record_requested(
		&mut self,
		record_key: KademliaKey,
		record_value: Vec<u8>,
		publisher: Option<PeerId>,
		expires: Option<std::time::Instant>,
	) -> Result<()> {
		let publisher = publisher.ok_or(Error::MissingPublisher)?;

		// Make sure we don't ever work with an outdated set of authorities
		// and that we do not update known_authorithies too often.
		let best_hash = self.client.best_hash().await?;
		if !self.known_authorities.contains_key(&record_key) &&
			self.authorities_queried_at
				.map(|authorities_queried_at| authorities_queried_at != best_hash)
				.unwrap_or(true)
		{
			let authorities = self
				.client
				.authorities(best_hash)
				.await
				.map_err(|e| Error::CallingRuntime(e.into()))?
				.into_iter()
				.collect::<Vec<_>>();

			self.known_authorities = authorities
				.into_iter()
				.map(|authority| (hash_authority_id(authority.as_ref()), authority))
				.collect::<HashMap<_, _>>();

			self.authorities_queried_at = Some(best_hash);
		}

		let authority_id =
			self.known_authorities.get(&record_key).ok_or(Error::UnknownAuthority)?;
		let signed_record =
			Self::check_record_signed_with_authority_id(record_value.as_slice(), authority_id)?;
		self.check_record_signed_with_network_key(
			&signed_record.record,
			signed_record.peer_signature,
			publisher,
			authority_id,
		)?;

		let records_creation_time: u128 =
			schema::AuthorityRecord::decode(signed_record.record.as_slice())
				.map_err(Error::DecodingProto)?
				.creation_time
				.map(|creation_time| {
					u128::decode(&mut &creation_time.timestamp[..]).unwrap_or_default()
				})
				.unwrap_or_default(); // 0 is a sane default for records that do not have creation time present.

		let current_record_info = self.last_known_records.get(&record_key);
		// If record creation time is older than the current record creation time,
		// we don't store it since we want to give higher priority to newer records.
		if let Some(current_record_info) = current_record_info {
			if records_creation_time < current_record_info.creation_time {
				debug!(
					target: LOG_TARGET,
					"Skip storing because record creation time {:?} is older than the current known record {:?}",
					records_creation_time,
					current_record_info.creation_time
				);
				return Ok(());
			}
		}

		self.network.store_record(record_key, record_value, Some(publisher), expires);
		Ok(())
	}

	fn check_record_signed_with_authority_id(
		record: &[u8],
		authority_id: &AuthorityId,
	) -> Result<schema::SignedAuthorityRecord> {
		let signed_record: schema::SignedAuthorityRecord =
			schema::SignedAuthorityRecord::decode(record).map_err(Error::DecodingProto)?;

		let auth_signature = AuthoritySignature::decode(&mut &signed_record.auth_signature[..])
			.map_err(Error::EncodingDecodingScale)?;

		if !AuthorityPair::verify(&auth_signature, &signed_record.record, &authority_id) {
			return Err(Error::VerifyingDhtPayload)
		}

		Ok(signed_record)
	}

	fn check_record_signed_with_network_key(
		&self,
		record: &Vec<u8>,
		peer_signature: Option<PeerSignature>,
		remote_peer_id: PeerId,
		authority_id: &AuthorityId,
	) -> Result<()> {
		if let Some(peer_signature) = peer_signature {
			match self.network.verify(
				remote_peer_id.into(),
				&peer_signature.public_key,
				&peer_signature.signature,
				record,
			) {
				Ok(true) => {},
				Ok(false) => return Err(Error::VerifyingDhtPayload),
				Err(error) => return Err(Error::ParsingLibp2pIdentity(error)),
			}
		} else if self.strict_record_validation {
			return Err(Error::MissingPeerIdSignature)
		} else {
			debug!(
				target: LOG_TARGET,
				"Received unsigned authority discovery record from {}", authority_id
			);
		}
		Ok(())
	}

	fn handle_dht_value_found_event(&mut self, peer_record: PeerRecord) -> Result<()> {
		// Ensure `values` is not empty and all its keys equal.
		let remote_key = peer_record.record.key.clone();

		let authority_id: AuthorityId =
			if let Some(authority_id) = self.in_flight_lookups.remove(&remote_key) {
				self.known_lookups.insert(remote_key.clone(), authority_id.clone());
				authority_id
			} else if let Some(authority_id) = self.known_lookups.get(&remote_key) {
				authority_id.clone()
			} else {
				return Err(Error::ReceivingUnexpectedRecord);
			};

		let local_peer_id = self.network.local_peer_id();

		let schema::SignedAuthorityRecord { record, peer_signature, .. } =
			Self::check_record_signed_with_authority_id(
				peer_record.record.value.as_slice(),
				&authority_id,
			)?;

		let authority_record =
			schema::AuthorityRecord::decode(record.as_slice()).map_err(Error::DecodingProto)?;

		let records_creation_time: u128 = authority_record
			.creation_time
			.as_ref()
			.map(|creation_time| {
				u128::decode(&mut &creation_time.timestamp[..]).unwrap_or_default()
			})
			.unwrap_or_default(); // 0 is a sane default for records that do not have creation time present.

		let addresses: Vec<Multiaddr> = authority_record
			.addresses
			.into_iter()
			.map(|a| a.try_into())
			.collect::<std::result::Result<_, _>>()
			.map_err(Error::ParsingMultiaddress)?;

		let get_peer_id = |a: &Multiaddr| match a.iter().last() {
			Some(multiaddr::Protocol::P2p(key)) => PeerId::from_multihash(key).ok(),
			_ => None,
		};

		// Ignore [`Multiaddr`]s without [`PeerId`] or with own addresses.
		let addresses: Vec<Multiaddr> = addresses
			.into_iter()
			.filter(|a| get_peer_id(&a).filter(|p| *p != local_peer_id).is_some())
			.collect();

		let remote_peer_id = single(addresses.iter().map(|a| get_peer_id(&a)))
			.map_err(|_| Error::ReceivingDhtValueFoundEventWithDifferentPeerIds)? // different peer_id in records
			.flatten()
			.ok_or(Error::ReceivingDhtValueFoundEventWithNoPeerIds)?; // no records with peer_id in them

		// At this point we know all the valid multiaddresses from the record, know that
		// each of them belong to the same PeerId, we just need to check if the record is
		// properly signed by the owner of the PeerId
		self.check_record_signed_with_network_key(
			&record,
			peer_signature,
			remote_peer_id,
			&authority_id,
		)?;

		let remote_addresses: Vec<Multiaddr> =
			addresses.into_iter().take(MAX_ADDRESSES_PER_AUTHORITY).collect();

		let answering_peer_id = peer_record.peer.map(|peer| peer.into());

		let addr_cache_needs_update = self.handle_new_record(
			&authority_id,
			remote_key.clone(),
			RecordInfo {
				creation_time: records_creation_time,
				peers_with_record: answering_peer_id.into_iter().collect(),
				record: peer_record.record,
			},
		);

		if !remote_addresses.is_empty() && addr_cache_needs_update {
			self.addr_cache.insert(authority_id, remote_addresses);
			if let Some(metrics) = &self.metrics {
				metrics
					.known_authorities_count
					.set(self.addr_cache.num_authority_ids().try_into().unwrap_or(std::u64::MAX));
			}
		}
		Ok(())
	}

	// Handles receiving a new DHT record for the authorithy.
	// Returns true if the record was new, false if the record was older than the current one.
	fn handle_new_record(
		&mut self,
		authority_id: &AuthorityId,
		kademlia_key: KademliaKey,
		new_record: RecordInfo,
	) -> bool {
		let current_record_info = self
			.last_known_records
			.entry(kademlia_key.clone())
			.or_insert_with(|| new_record.clone());

		if new_record.creation_time > current_record_info.creation_time {
			let peers_that_need_updating = current_record_info.peers_with_record.clone();
			self.network.put_record_to(
				new_record.record.clone(),
				peers_that_need_updating.clone(),
				// If this is empty it means we received the answer from our node local
				// storage, so we need to update that as well.
				current_record_info.peers_with_record.is_empty(),
			);
			debug!(
					target: LOG_TARGET,
					"Found a newer record for {:?} new record creation time {:?} old record creation time {:?}",
					authority_id, new_record.creation_time, current_record_info.creation_time
			);
			self.last_known_records.insert(kademlia_key, new_record);
			return true
		}

		if new_record.creation_time == current_record_info.creation_time {
			// Same record just update in case this is a record from old nodes that don't have
			// timestamp.
			debug!(
					target: LOG_TARGET,
					"Found same record for {:?} record creation time {:?}",
					authority_id, new_record.creation_time
			);
			if current_record_info.peers_with_record.len() + new_record.peers_with_record.len() <=
				DEFAULT_KADEMLIA_REPLICATION_FACTOR
			{
				current_record_info.peers_with_record.extend(new_record.peers_with_record);
			}
			return true
		}

		debug!(
				target: LOG_TARGET,
				"Found old record for {:?} received record creation time {:?} current record creation time {:?}",
				authority_id, new_record.creation_time, current_record_info.creation_time,
		);
		self.network.put_record_to(
			current_record_info.record.clone(),
			new_record.peers_with_record.clone(),
			// If this is empty it means we received the answer from our node local
			// storage, so we need to update that as well.
			new_record.peers_with_record.is_empty(),
		);
		return false
	}

	/// Retrieve our public keys within the current and next authority set.
	// A node might have multiple authority discovery keys within its keystore, e.g. an old one and
	// one for the upcoming session. In addition it could be participating in the current and (/ or)
	// next authority set with two keys. The function does not return all of the local authority
	// discovery public keys, but only the ones intersecting with the current or next authority set.
	async fn get_own_public_keys_within_authority_set(
		key_store: KeystorePtr,
		client: &Client,
	) -> Result<HashSet<AuthorityId>> {
		let local_pub_keys = key_store
			.sr25519_public_keys(key_types::AUTHORITY_DISCOVERY)
			.into_iter()
			.collect::<HashSet<_>>();

		let best_hash = client.best_hash().await?;
		let authorities = client
			.authorities(best_hash)
			.await
			.map_err(|e| Error::CallingRuntime(e.into()))?
			.into_iter()
			.map(Into::into)
			.collect::<HashSet<_>>();

		let intersection =
			local_pub_keys.intersection(&authorities).cloned().map(Into::into).collect();

		Ok(intersection)
	}
}

/// NetworkProvider provides [`Worker`] with all necessary hooks into the
/// underlying Substrate networking. Using this trait abstraction instead of
/// `sc_network::NetworkService` directly is necessary to unit test [`Worker`].
pub trait NetworkProvider:
	NetworkDHTProvider + NetworkStateInfo + NetworkSigner + Send + Sync
{
}

impl<T> NetworkProvider for T where
	T: NetworkDHTProvider + NetworkStateInfo + NetworkSigner + Send + Sync
{
}

fn hash_authority_id(id: &[u8]) -> KademliaKey {
	KademliaKey::new(&Code::Sha2_256.digest(id).digest())
}

// Makes sure all values are the same and returns it
//
// Returns Err(_) if not all values are equal. Returns Ok(None) if there are
// no values.
fn single<T>(values: impl IntoIterator<Item = T>) -> std::result::Result<Option<T>, ()>
where
	T: PartialEq<T>,
{
	values.into_iter().try_fold(None, |acc, item| match acc {
		None => Ok(Some(item)),
		Some(ref prev) if *prev != item => Err(()),
		Some(x) => Ok(Some(x)),
	})
}

fn serialize_addresses(addresses: impl Iterator<Item = Multiaddr>) -> Vec<Vec<u8>> {
	addresses.map(|a| a.to_vec()).collect()
}

fn build_creation_time() -> schema::TimestampInfo {
	let creation_time = SystemTime::now()
		.duration_since(UNIX_EPOCH)
		.map(|time| time.as_nanos())
		.unwrap_or_default();
	schema::TimestampInfo { timestamp: creation_time.encode() }
}

fn serialize_authority_record(
	addresses: Vec<Vec<u8>>,
	creation_time: Option<schema::TimestampInfo>,
) -> Result<Vec<u8>> {
	let mut serialized_record = vec![];

	schema::AuthorityRecord { addresses, creation_time }
		.encode(&mut serialized_record)
		.map_err(Error::EncodingProto)?;
	Ok(serialized_record)
}

fn sign_record_with_peer_id(
	serialized_record: &[u8],
	network: &impl NetworkSigner,
) -> Result<schema::PeerSignature> {
	let signature = network
		.sign_with_local_identity(serialized_record.to_vec())
		.map_err(|e| Error::CannotSign(format!("{} (network packet)", e)))?;
	let public_key = signature.public_key.encode_protobuf();
	let signature = signature.bytes;
	Ok(schema::PeerSignature { signature, public_key })
}

fn sign_record_with_authority_ids(
	serialized_record: Vec<u8>,
	peer_signature: Option<schema::PeerSignature>,
	key_store: &dyn Keystore,
	keys: Vec<AuthorityId>,
) -> Result<Vec<(KademliaKey, Vec<u8>)>> {
	let mut result = Vec::with_capacity(keys.len());

	for key in keys.iter() {
		let auth_signature = key_store
			.sr25519_sign(key_types::AUTHORITY_DISCOVERY, key.as_ref(), &serialized_record)
			.map_err(|e| Error::CannotSign(format!("{}. Key: {:?}", e, key)))?
			.ok_or_else(|| {
				Error::CannotSign(format!("Could not find key in keystore. Key: {:?}", key))
			})?;

		// Scale encode
		let auth_signature = auth_signature.encode();
		let signed_record = schema::SignedAuthorityRecord {
			record: serialized_record.clone(),
			auth_signature,
			peer_signature: peer_signature.clone(),
		}
		.encode_to_vec();

		result.push((hash_authority_id(key.as_slice()), signed_record));
	}

	Ok(result)
}

/// Prometheus metrics for a [`Worker`].
#[derive(Clone)]
pub(crate) struct Metrics {
	publish: Counter<U64>,
	amount_addresses_last_published: Gauge<U64>,
	requests: Counter<U64>,
	requests_pending: Gauge<U64>,
	dht_event_received: CounterVec<U64>,
	handle_value_found_event_failure: Counter<U64>,
	known_authorities_count: Gauge<U64>,
}

impl Metrics {
	pub(crate) fn register(registry: &prometheus_endpoint::Registry) -> Result<Self> {
		Ok(Self {
			publish: register(
				Counter::new(
					"substrate_authority_discovery_times_published_total",
					"Number of times authority discovery has published external addresses.",
				)?,
				registry,
			)?,
			amount_addresses_last_published: register(
				Gauge::new(
					"substrate_authority_discovery_amount_external_addresses_last_published",
					"Number of external addresses published when authority discovery last \
					 published addresses.",
				)?,
				registry,
			)?,
			requests: register(
				Counter::new(
					"substrate_authority_discovery_authority_addresses_requested_total",
					"Number of times authority discovery has requested external addresses of a \
					 single authority.",
				)?,
				registry,
			)?,
			requests_pending: register(
				Gauge::new(
					"substrate_authority_discovery_authority_address_requests_pending",
					"Number of pending authority address requests.",
				)?,
				registry,
			)?,
			dht_event_received: register(
				CounterVec::new(
					Opts::new(
						"substrate_authority_discovery_dht_event_received",
						"Number of dht events received by authority discovery.",
					),
					&["name"],
				)?,
				registry,
			)?,
			handle_value_found_event_failure: register(
				Counter::new(
					"substrate_authority_discovery_handle_value_found_event_failure",
					"Number of times handling a dht value found event failed.",
				)?,
				registry,
			)?,
			known_authorities_count: register(
				Gauge::new(
					"substrate_authority_discovery_known_authorities_count",
					"Number of authorities known by authority discovery.",
				)?,
				registry,
			)?,
		})
	}
}

// Helper functions for unit testing.
#[cfg(test)]
impl<Block: BlockT, Client, DhtEventStream> Worker<Client, Block, DhtEventStream> {
	pub(crate) fn inject_addresses(&mut self, authority: AuthorityId, addresses: Vec<Multiaddr>) {
		self.addr_cache.insert(authority, addresses);
	}
}