Merge branch 'development' into haiko/remove-ecs-gha

.github/workflows/build.yml

@@ -103,7 +103,7 @@ jobs:
       - name: Generate coverage report
         run: |
           go test ./... -coverprofile=coverage.out -covermode=atomic -timeout=20m
-      - uses: codecov/codecov-action@v5.1.2
+      - uses: codecov/codecov-action@v5.3.1
         with:
           files: ./coverage.out
           flags: unit-tests

.github/workflows/unit-tests.yml

@@ -89,7 +89,7 @@ jobs:
       - name: Test - Race
         run: make test-using-race-detector
 
-      - uses: codecov/codecov-action@v5.1.2
+      - uses: codecov/codecov-action@v5.3.1
         with:
           if_ci_failed: success
           informational: true

README.md

@@ -35,6 +35,9 @@
 >
 > The Gossamer Polkadot Host is pre-production software [2022-12-01]
 
+>Developed by [ChainSafe](https://chainsafe.io) and funded by a Web3 Foundation grant and Polkadot [Treasury proposals](https://polkadot.polkassembly.io/treasury/808?tab=description).
+>
+
 Gossamer is a [Golang](https://go.dev/) implementation of the
 [Polkadot Host](https://wiki.polkadot.network/docs/learn-polkadot-host): an
 execution environment for the Polkadot runtime, which is materialized as a Web
@@ -43,6 +46,7 @@ Polkadot Host must orchestrate a number of interrelated services, such as
 [networking](dot/network/README.md), block production, block finalization, a
 JSON-RPC server, [and more](cmd/gossamer/README.md#client-components).
 
+
 ## Getting Started
 
 To get started with Gossamer, follow the steps below to build the source code

chain/kusama/defaults.go

@@ -25,6 +25,7 @@ func DefaultConfig() *cfg.Config {
 	config.Core.GrandpaAuthority = false
 	config.Core.Role = 1
 	config.Network.NoMDNS = false
+	config.Core.Sync = "full"
 
 	return config
 }

chain/paseo/defaults.go

@@ -24,6 +24,7 @@ func DefaultConfig() *cfg.Config {
 	config.Core.GrandpaAuthority = false
 	config.Core.Role = 1
 	config.Network.NoMDNS = false
+	config.Core.Sync = "full"
 
 	return config
 }

chain/polkadot/defaults.go

@@ -24,6 +24,7 @@ func DefaultConfig() *cfg.Config {
 	config.Core.GrandpaAuthority = false
 	config.Core.Role = 1
 	config.Network.NoMDNS = false
+	config.Core.Sync = "full"
 
 	return config
 }

chain/westend-dev/defaults.go

@@ -24,6 +24,7 @@ func DefaultConfig() *cfg.Config {
 	config.RPC.UnsafeRPC = true
 	config.RPC.WSExternal = true
 	config.RPC.UnsafeWSExternal = true
+	config.Core.Sync = "full"
 
 	return config
 }

chain/westend-local/default.go

@@ -29,6 +29,7 @@ func DefaultConfig() *cfg.Config {
 	config.RPC.UnsafeRPC = true
 	config.RPC.WSExternal = true
 	config.RPC.UnsafeWSExternal = true
+	config.Core.Sync = "full"
 
 	return config
 }

chain/westend/defaults.go

@@ -24,6 +24,7 @@ func DefaultConfig() *cfg.Config {
 	config.Core.GrandpaAuthority = false
 	config.Core.Role = 1
 	config.Network.NoMDNS = false
+	config.Core.Sync = "full"
 
 	return config
 }

cmd/gossamer/commands/root.go

@@ -529,6 +529,14 @@ func addCoreFlags(cmd *cobra.Command) error {
 		return fmt.Errorf("failed to add --grandpa-interval flag: %s", err)
 	}
 
+	if err := addStringFlagBindViper(cmd,
+		"sync",
+		config.Core.Sync,
+		"sync mode [warp | full]",
+		"core.sync"); err != nil {
+		return fmt.Errorf("failed to add --sync flag: %s", err)
+	}
+
 	return nil
 }
 

config/config.go

@@ -61,6 +61,9 @@ const (
 	DefaultSystemName = "Gossamer"
 	// DefaultSystemVersion is the default system version
 	DefaultSystemVersion = "0.0.0"
+
+	// DefaultSyncMode is the default block sync mode
+	DefaultSyncMode = "full"
 )
 
 // DefaultRPCModules the default RPC modules
@@ -188,6 +191,7 @@ type CoreConfig struct {
 	GrandpaAuthority bool               `mapstructure:"grandpa-authority"`
 	WasmInterpreter  string             `mapstructure:"wasm-interpreter,omitempty"`
 	GrandpaInterval  time.Duration      `mapstructure:"grandpa-interval,omitempty"`
+	Sync             string             `mapstructure:"sync,omitempty"`
 }
 
 // StateConfig contains the configuration for the state.
@@ -363,6 +367,7 @@ func DefaultConfig() *Config {
 			GrandpaAuthority: true,
 			WasmInterpreter:  DefaultWasmInterpreter,
 			GrandpaInterval:  DefaultDiscoveryInterval,
+			Sync:             DefaultSyncMode,
 		},
 		Network: &NetworkConfig{
 			Port:              DefaultNetworkPort,
@@ -444,6 +449,7 @@ func DefaultConfigFromSpec(nodeSpec *genesis.Genesis) *Config {
 			GrandpaAuthority: true,
 			WasmInterpreter:  DefaultWasmInterpreter,
 			GrandpaInterval:  DefaultDiscoveryInterval,
+			Sync:             DefaultSyncMode,
 		},
 		Network: &NetworkConfig{
 			Port:              DefaultNetworkPort,
@@ -525,6 +531,7 @@ func Copy(c *Config) Config {
 			GrandpaAuthority: c.Core.GrandpaAuthority,
 			WasmInterpreter:  c.Core.WasmInterpreter,
 			GrandpaInterval:  c.Core.GrandpaInterval,
+			Sync:             c.Core.Sync,
 		},
 		Network: &NetworkConfig{
 			Port:              c.Network.Port,

docs/docs/design/authority-discovery.md

@@ -0,0 +1,39 @@
+---
+layout: default
+title: Authority Discovery Overview
+permalink: /design/authority-discovery/
+---
+
+# Authority Discovery
+
+## Questions
+
+### Is the authority discovery implemented on Kademlia for a custom peer discovery?
+
+No, it uses Kademlia capabilities however it adds a custom handlers for DHT Events, like: DhtEvent::ValueFound
+- DhtEvent::ValueNotFound
+- DhtEvent::ValuePut
+- DhtEvent::ValuePutFailed
+- DhtEvent::PutRecordRequest
+
+These events are produced by the Discovery layer (Kademlia) and propagated to the `authority-discovery` that handles each of them. Specifically for `DhtEvent::PutRecordRequest` it does a validation step and then store the record. 
+
+What happens is that the `authority-discovery` is the actual protocol implementation for discovery in substrate/polkadot given that it produces and store the records that are available in the DHT for other peers to query, validate and use.
+
+### What are the functionalities of the authority discovery, beyond discovery new peers based on its authority ids?
+
+- Publish the node external address, more specifically, create the authority discovery record, sign it and distributing through DHT allowing other nodes to query it.
+- Handle DHT events, validating incoming records and caching other authorities address.
+
+### What subsystems uses the authority discovery mechanism? Describe the usages.
+
+- Availability Distribution
+- Availability Recovery
+- Gossip Support
+
+### Check if we can currently extend the authority discovery protocol can to golang kademlia p2p library
+
+Yes, we pretty much can implement the protocol. 
+- We are able to encode the records using the same [proto schema used by substrate/polkadot](https://github.com/paritytech/polkadot-sdk/blob/master/substrate/client/authority-discovery/src/worker/schema/dht-v3.proto) - We can query for specific keys in DHT.
+- What can write a custom validator to validate incoming records.
+- golang-libp2p does not works in the same way rust-libp2p works by enabling us to receive those DHT events but we can achieve the same protocol implementation using what the library provide to us.

docs/docs/design/gossip-support.md

@@ -0,0 +1,36 @@
+---
+layout: default
+title: Gossip Support
+permalink: /design/gossip-support/
+---
+
+# Gossip Support
+
+Gossip support is a short subsystem that cares about handling active leaves and network bridge update events.
+
+## Functionalities
+
+- **Check connectivity**: The subsystem contains a timer that triggers every 600 seconds (10 minutes) a connectivity check that is a simple verification on the amount of authorities connected against the actual set of authorities, if the percentage of connected authorities is bellow 90% the node will issue WARN logs.
+
+- **Handle Active Leaves**: 
+  - Given the current values in the state, **last failure** and **last connection request**, the subsystem should force new requests or re-resolve authorities
+  - Determine if the current session index has changed, if so determine relevant validators and issue connection request. The subsystem should change the **last sesison index** value only if it is possible to retrieve the new session info.
+  - If we notice that a new session is starting we should update our authority ids cache and send a `NetworkBridgeRxMessage::UpdatedAuthorityIds` message with informations about the new set of authorities.
+
+- **Handle Network Bridge Update**:
+  - In the above point, when handling active leaves, we issue connection requests and whenever a new connection is stablished or a disconnection happens for some reason. The subsytems listen on any network bridge update about `NetworkBridgeEvent::PeerConnected` and `NetworkBridgeEvent::PeerDisconnected`, other updates are not important.
+
+## State
+
+The subsystem state is composed of:
+
+- Keystore
+- Last Session Index
+- **Last Failure**: last time we could not resolve a third of the authorities
+- **Last Connection Request**: is possible that validators change their peer IDs in during a session. If that happens we will reconnect to them in the best case after a session, so we need to try more often to resolve peers and reconnect to them. We cannot detect changes faster them Authority Discovery DHT queries.
+- **Failure Start**: first time the node cannot reach the connectivity threshold
+- **Resolved Authorities**: A map of resolved authority IDs to a set of multiaddr, basically we were able to find the addresses for authorities, not mean we are actually connected to them.
+- **Connected Authorities**: A map of actual authorities ID's connected to their PeerId
+- **Connected Peers**: A map of PeerId to authority IDs for fast lookup
+- **Authority Discovery**: authority discovery service, allows us to resolve authorities addresses.
+

docs/docs/design/statement-distribution.md

@@ -1,3 +1,9 @@
+---
+layout: default
+title: Statement Distribution
+permalink: /design/statement-distribution/
+---
+
 # Statement Distribution
 
 This subsystem is responsible for distributing signed statements that we have generated and forwarding statements generated by our peers.

docs/mkdocs.yml

@@ -73,6 +73,10 @@ nav:
     - Availability Subsystems: ./design/availability-subsystems.md
     - Prospective Parachains: ./design/prospective-parachains.md
     - Prospective Parachains - Fragment Chain: ./design/fragment-chain.md
+    - Authority Discovery: ./design/authority-discovery.md
+    - Statement Distribution: ./design/statement-distribution.md
+    - Grid Topology: ./design/grid-topology.md
+    - Gossip Support: ./design/gossip-support.md
   - Contributing: 
     - Overview: ./contributing.md
 extra_css:

dot/network/host_integration_test.go

@@ -7,6 +7,7 @@ package network
 
 import (
 	"fmt"
+	"strings"
 	"testing"
 	"time"
 
@@ -396,12 +397,21 @@ func Test_PeerSupportsProtocol(t *testing.T) {
 	}
 	require.NoError(t, err)
 
+	genesisHash := nodeA.blockState.GenesisHash().String()
+	genesisHash = strings.TrimPrefix(genesisHash, "0x")
+	fullSyncProtocolId := fmt.Sprintf("/%s%s", genesisHash, SyncID)
+	warpSyncProtocolId := fmt.Sprintf("/%s%s", genesisHash, WarpSyncID)
+
 	tests := []struct {
 		protocol protocol.ID
 		expect   bool
 	}{
 		{
-			protocol: protocol.ID("/gossamer/test/0/sync/2"),
+			protocol: protocol.ID(fullSyncProtocolId),
+			expect:   true,
+		},
+		{
+			protocol: protocol.ID(warpSyncProtocolId),
 			expect:   true,
 		},
 		{

dot/network/messages/warp_sync.go

@@ -15,6 +15,12 @@ type WarpProofRequest struct {
 	Begin common.Hash
 }
 
+func NewWarpProofRequest(from common.Hash) *WarpProofRequest {
+	return &WarpProofRequest{
+		Begin: from,
+	}
+}
+
 // Decode decodes the message into a WarpProofRequest
 func (wpr *WarpProofRequest) Decode(in []byte) error {
 	return scale.Unmarshal(in, wpr)

dot/network/service.go

@@ -259,11 +259,14 @@ func (s *Service) Start() error {
 		s.ctx, s.cancel = context.WithCancel(context.Background())
 	}
 
-	genesisHashProtocolId := protocol.ID(s.cfg.BlockState.GenesisHash().String())
+	genesisHash := s.blockState.GenesisHash().String()
+	genesisHash = strings.TrimPrefix(genesisHash, "0x")
+	fullSyncProtocolId := fmt.Sprintf("/%s%s", genesisHash, SyncID)
+	warpSyncProtocolId := fmt.Sprintf("/%s%s", genesisHash, WarpSyncID)
 
-	s.host.registerStreamHandler(s.host.protocolID+SyncID, s.handleSyncStream)
+	s.host.registerStreamHandler(protocol.ID(fullSyncProtocolId), s.handleSyncStream)
 	s.host.registerStreamHandler(s.host.protocolID+lightID, s.handleLightStream)
-	s.host.registerStreamHandler(genesisHashProtocolId+WarpSyncID, s.handleWarpSyncStream)
+	s.host.registerStreamHandler(protocol.ID(warpSyncProtocolId), s.handleWarpSyncStream)
 
 	// register block announce protocol
 	err := s.RegisterNotificationsProtocol(
@@ -622,13 +625,16 @@ func (s *Service) SendMessage(to peer.ID, msg NotificationsMessage) error {
 func (s *Service) GetRequestResponseProtocol(subprotocol string, requestTimeout time.Duration,
 	maxResponseSize uint64) *RequestResponseProtocol {
 
-	protocolID := s.host.protocolID + protocol.ID(subprotocol)
+	genesisHash := s.blockState.GenesisHash().String()
+	genesisHash = strings.TrimPrefix(genesisHash, "0x")
+	protocolId := fmt.Sprintf("/%s%s", genesisHash, subprotocol)
+
 	return &RequestResponseProtocol{
 		ctx:             s.ctx,
 		host:            s.host,
 		requestTimeout:  requestTimeout,
 		maxResponseSize: maxResponseSize,
-		protocolID:      protocolID,
+		protocolID:      protocol.ID(protocolId),
 		responseBuf:     make([]byte, maxResponseSize),
 		responseBufMu:   sync.Mutex{},
 	}

dot/network/warp_sync.go

@@ -8,23 +8,16 @@ import (
 	"fmt"
 
 	"github.com/ChainSafe/gossamer/dot/network/messages"
-	"github.com/ChainSafe/gossamer/dot/types"
 	"github.com/ChainSafe/gossamer/internal/primitives/consensus/grandpa"
 	primitives "github.com/ChainSafe/gossamer/internal/primitives/consensus/grandpa"
 	"github.com/ChainSafe/gossamer/lib/common"
+	"github.com/ChainSafe/gossamer/lib/grandpa/warpsync"
 	libp2pnetwork "github.com/libp2p/go-libp2p/core/network"
 	"github.com/libp2p/go-libp2p/core/peer"
 )
 
 const MaxAllowedSameRequestPerPeer = 5
 
-type WarpSyncVerificationResult struct {
-	SetId         grandpa.SetID
-	AuthorityList primitives.AuthorityList
-	Header        types.Header
-	Completed     bool
-}
-
 // WarpSyncProvider is an interface for generating warp sync proofs
 type WarpSyncProvider interface {
 	// Generate proof starting at given block hash. The proof is accumulated until maximum proof
@@ -34,7 +27,7 @@ type WarpSyncProvider interface {
 		encodedProof []byte,
 		setId grandpa.SetID,
 		authorities primitives.AuthorityList,
-	) (*WarpSyncVerificationResult, error)
+	) (*warpsync.WarpSyncVerificationResult, error)
 }
 
 func (s *Service) handleWarpSyncRequest(req messages.WarpProofRequest) ([]byte, error) {

dot/peerset/constants.go

@@ -76,4 +76,14 @@ const (
 	// SameBlockSyncRequest used when a peer send us more than the max number of the same request.
 	SameBlockSyncRequest       Reputation = math.MinInt32
 	SameBlockSyncRequestReason            = "same block sync request"
+
+	// UnexpectedResponseValue is used when peer send an unexpected response.
+	UnexpectedResponseValue Reputation = -(1 << 29)
+	// UnexpectedResponseReason is used when peer send an unexpected response.
+	UnexpectedResponseReason = "Unexpected response"
+
+	// BadWarpProofValue is used when peer send invalid warp sync proof.
+	BadWarpProofValue Reputation = -(1 << 29)
+	// BadWarpProofReason is used when peer send invalid warp sync proof.
+	BadWarpProofReason = "Bad warp proof"
 )