Browse Source

Merge pull request #653 from tendermint/improvement/peer_interface

peer interface
pull/657/head
Ethan Buchman 7 years ago
committed by GitHub
parent
commit
126f63c0a2
15 changed files with 195 additions and 148 deletions
  1. +6
    -6
      blockchain/reactor.go
  2. +6
    -6
      consensus/byzantine_test.go
  3. +2
    -2
      consensus/common_test.go
  4. +30
    -23
      consensus/reactor.go
  5. +3
    -3
      mempool/reactor.go
  6. +71
    -33
      p2p/peer.go
  7. +16
    -16
      p2p/peer_set.go
  8. +11
    -11
      p2p/peer_set_test.go
  9. +3
    -3
      p2p/peer_test.go
  10. +9
    -9
      p2p/pex_reactor.go
  11. +7
    -6
      p2p/pex_reactor_test.go
  12. +21
    -20
      p2p/switch.go
  13. +6
    -6
      p2p/switch_test.go
  14. +2
    -2
      rpc/core/consensus.go
  15. +2
    -2
      rpc/core/net.go

+ 6
- 6
blockchain/reactor.go View File

@ -111,19 +111,19 @@ func (bcR *BlockchainReactor) GetChannels() []*p2p.ChannelDescriptor {
}
// AddPeer implements Reactor by sending our state to peer.
func (bcR *BlockchainReactor) AddPeer(peer *p2p.Peer) {
func (bcR *BlockchainReactor) AddPeer(peer p2p.Peer) {
if !peer.Send(BlockchainChannel, struct{ BlockchainMessage }{&bcStatusResponseMessage{bcR.store.Height()}}) {
// doing nothing, will try later in `poolRoutine`
}
}
// RemovePeer implements Reactor by removing peer from the pool.
func (bcR *BlockchainReactor) RemovePeer(peer *p2p.Peer, reason interface{}) {
bcR.pool.RemovePeer(peer.Key)
func (bcR *BlockchainReactor) RemovePeer(peer p2p.Peer, reason interface{}) {
bcR.pool.RemovePeer(peer.Key())
}
// Receive implements Reactor by handling 4 types of messages (look below).
func (bcR *BlockchainReactor) Receive(chID byte, src *p2p.Peer, msgBytes []byte) {
func (bcR *BlockchainReactor) Receive(chID byte, src p2p.Peer, msgBytes []byte) {
_, msg, err := DecodeMessage(msgBytes)
if err != nil {
bcR.Logger.Error("Error decoding message", "err", err)
@ -148,7 +148,7 @@ func (bcR *BlockchainReactor) Receive(chID byte, src *p2p.Peer, msgBytes []byte)
}
case *bcBlockResponseMessage:
// Got a block.
bcR.pool.AddBlock(src.Key, msg.Block, len(msgBytes))
bcR.pool.AddBlock(src.Key(), msg.Block, len(msgBytes))
case *bcStatusRequestMessage:
// Send peer our state.
queued := src.TrySend(BlockchainChannel, struct{ BlockchainMessage }{&bcStatusResponseMessage{bcR.store.Height()}})
@ -157,7 +157,7 @@ func (bcR *BlockchainReactor) Receive(chID byte, src *p2p.Peer, msgBytes []byte)
}
case *bcStatusResponseMessage:
// Got a peer status. Unverified.
bcR.pool.SetPeerHeight(src.Key, msg.Height)
bcR.pool.SetPeerHeight(src.Key(), msg.Height)
default:
bcR.Logger.Error(cmn.Fmt("Unknown message type %v", reflect.TypeOf(msg)))
}


+ 6
- 6
consensus/byzantine_test.go View File

@ -188,7 +188,7 @@ func byzantineDecideProposalFunc(t *testing.T, height, round int, cs *ConsensusS
}
}
func sendProposalAndParts(height, round int, cs *ConsensusState, peer *p2p.Peer, proposal *types.Proposal, blockHash []byte, parts *types.PartSet) {
func sendProposalAndParts(height, round int, cs *ConsensusState, peer p2p.Peer, proposal *types.Proposal, blockHash []byte, parts *types.PartSet) {
// proposal
msg := &ProposalMessage{Proposal: proposal}
peer.Send(DataChannel, struct{ ConsensusMessage }{msg})
@ -231,14 +231,14 @@ func NewByzantineReactor(conR *ConsensusReactor) *ByzantineReactor {
func (br *ByzantineReactor) SetSwitch(s *p2p.Switch) { br.reactor.SetSwitch(s) }
func (br *ByzantineReactor) GetChannels() []*p2p.ChannelDescriptor { return br.reactor.GetChannels() }
func (br *ByzantineReactor) AddPeer(peer *p2p.Peer) {
func (br *ByzantineReactor) AddPeer(peer p2p.Peer) {
if !br.reactor.IsRunning() {
return
}
// Create peerState for peer
peerState := NewPeerState(peer)
peer.Data.Set(types.PeerStateKey, peerState)
peerState := NewPeerState(peer).SetLogger(br.reactor.Logger)
peer.Set(types.PeerStateKey, peerState)
// Send our state to peer.
// If we're fast_syncing, broadcast a RoundStepMessage later upon SwitchToConsensus().
@ -246,10 +246,10 @@ func (br *ByzantineReactor) AddPeer(peer *p2p.Peer) {
br.reactor.sendNewRoundStepMessages(peer)
}
}
func (br *ByzantineReactor) RemovePeer(peer *p2p.Peer, reason interface{}) {
func (br *ByzantineReactor) RemovePeer(peer p2p.Peer, reason interface{}) {
br.reactor.RemovePeer(peer, reason)
}
func (br *ByzantineReactor) Receive(chID byte, peer *p2p.Peer, msgBytes []byte) {
func (br *ByzantineReactor) Receive(chID byte, peer p2p.Peer, msgBytes []byte) {
br.reactor.Receive(chID, peer, msgBytes)
}


+ 2
- 2
consensus/common_test.go View File

@ -380,9 +380,9 @@ func randConsensusNetWithPeers(nValidators, nPeers int, testName string, tickerF
return css
}
func getSwitchIndex(switches []*p2p.Switch, peer *p2p.Peer) int {
func getSwitchIndex(switches []*p2p.Switch, peer p2p.Peer) int {
for i, s := range switches {
if bytes.Equal(peer.NodeInfo.PubKey.Address(), s.NodeInfo().PubKey.Address()) {
if bytes.Equal(peer.NodeInfo().PubKey.Address(), s.NodeInfo().PubKey.Address()) {
return i
}
}


+ 30
- 23
consensus/reactor.go View File

@ -120,14 +120,14 @@ func (conR *ConsensusReactor) GetChannels() []*p2p.ChannelDescriptor {
}
// AddPeer implements Reactor
func (conR *ConsensusReactor) AddPeer(peer *p2p.Peer) {
func (conR *ConsensusReactor) AddPeer(peer p2p.Peer) {
if !conR.IsRunning() {
return
}
// Create peerState for peer
peerState := NewPeerState(peer)
peer.Data.Set(types.PeerStateKey, peerState)
peerState := NewPeerState(peer).SetLogger(conR.Logger)
peer.Set(types.PeerStateKey, peerState)
// Begin routines for this peer.
go conR.gossipDataRoutine(peer, peerState)
@ -142,12 +142,12 @@ func (conR *ConsensusReactor) AddPeer(peer *p2p.Peer) {
}
// RemovePeer implements Reactor
func (conR *ConsensusReactor) RemovePeer(peer *p2p.Peer, reason interface{}) {
func (conR *ConsensusReactor) RemovePeer(peer p2p.Peer, reason interface{}) {
if !conR.IsRunning() {
return
}
// TODO
//peer.Data.Get(PeerStateKey).(*PeerState).Disconnect()
//peer.Get(PeerStateKey).(*PeerState).Disconnect()
}
// Receive implements Reactor
@ -156,7 +156,7 @@ func (conR *ConsensusReactor) RemovePeer(peer *p2p.Peer, reason interface{}) {
// Peer state updates can happen in parallel, but processing of
// proposals, block parts, and votes are ordered by the receiveRoutine
// NOTE: blocks on consensus state for proposals, block parts, and votes
func (conR *ConsensusReactor) Receive(chID byte, src *p2p.Peer, msgBytes []byte) {
func (conR *ConsensusReactor) Receive(chID byte, src p2p.Peer, msgBytes []byte) {
if !conR.IsRunning() {
conR.Logger.Debug("Receive", "src", src, "chId", chID, "bytes", msgBytes)
return
@ -171,7 +171,7 @@ func (conR *ConsensusReactor) Receive(chID byte, src *p2p.Peer, msgBytes []byte)
conR.Logger.Debug("Receive", "src", src, "chId", chID, "msg", msg)
// Get peer states
ps := src.Data.Get(types.PeerStateKey).(*PeerState)
ps := src.Get(types.PeerStateKey).(*PeerState)
switch chID {
case StateChannel:
@ -191,7 +191,7 @@ func (conR *ConsensusReactor) Receive(chID byte, src *p2p.Peer, msgBytes []byte)
return
}
// Peer claims to have a maj23 for some BlockID at H,R,S,
votes.SetPeerMaj23(msg.Round, msg.Type, ps.Peer.Key, msg.BlockID)
votes.SetPeerMaj23(msg.Round, msg.Type, ps.Peer.Key(), msg.BlockID)
// Respond with a VoteSetBitsMessage showing which votes we have.
// (and consequently shows which we don't have)
var ourVotes *cmn.BitArray
@ -228,12 +228,12 @@ func (conR *ConsensusReactor) Receive(chID byte, src *p2p.Peer, msgBytes []byte)
switch msg := msg.(type) {
case *ProposalMessage:
ps.SetHasProposal(msg.Proposal)
conR.conS.peerMsgQueue <- msgInfo{msg, src.Key}
conR.conS.peerMsgQueue <- msgInfo{msg, src.Key()}
case *ProposalPOLMessage:
ps.ApplyProposalPOLMessage(msg)
case *BlockPartMessage:
ps.SetHasProposalBlockPart(msg.Height, msg.Round, msg.Part.Index)
conR.conS.peerMsgQueue <- msgInfo{msg, src.Key}
conR.conS.peerMsgQueue <- msgInfo{msg, src.Key()}
default:
conR.Logger.Error(cmn.Fmt("Unknown message type %v", reflect.TypeOf(msg)))
}
@ -253,7 +253,7 @@ func (conR *ConsensusReactor) Receive(chID byte, src *p2p.Peer, msgBytes []byte)
ps.EnsureVoteBitArrays(height-1, lastCommitSize)
ps.SetHasVote(msg.Vote)
cs.peerMsgQueue <- msgInfo{msg, src.Key}
cs.peerMsgQueue <- msgInfo{msg, src.Key()}
default:
// don't punish (leave room for soft upgrades)
@ -367,7 +367,7 @@ func (conR *ConsensusReactor) broadcastHasVoteMessage(vote *types.Vote) {
/*
// TODO: Make this broadcast more selective.
for _, peer := range conR.Switch.Peers().List() {
ps := peer.Data.Get(PeerStateKey).(*PeerState)
ps := peer.Get(PeerStateKey).(*PeerState)
prs := ps.GetRoundState()
if prs.Height == vote.Height {
// TODO: Also filter on round?
@ -399,7 +399,7 @@ func makeRoundStepMessages(rs *RoundState) (nrsMsg *NewRoundStepMessage, csMsg *
return
}
func (conR *ConsensusReactor) sendNewRoundStepMessages(peer *p2p.Peer) {
func (conR *ConsensusReactor) sendNewRoundStepMessages(peer p2p.Peer) {
rs := conR.conS.GetRoundState()
nrsMsg, csMsg := makeRoundStepMessages(rs)
if nrsMsg != nil {
@ -410,7 +410,7 @@ func (conR *ConsensusReactor) sendNewRoundStepMessages(peer *p2p.Peer) {
}
}
func (conR *ConsensusReactor) gossipDataRoutine(peer *p2p.Peer, ps *PeerState) {
func (conR *ConsensusReactor) gossipDataRoutine(peer p2p.Peer, ps *PeerState) {
logger := conR.Logger.With("peer", peer)
OUTER_LOOP:
@ -492,7 +492,7 @@ OUTER_LOOP:
}
func (conR *ConsensusReactor) gossipDataForCatchup(logger log.Logger, rs *RoundState,
prs *PeerRoundState, ps *PeerState, peer *p2p.Peer) {
prs *PeerRoundState, ps *PeerState, peer p2p.Peer) {
if index, ok := prs.ProposalBlockParts.Not().PickRandom(); ok {
// Ensure that the peer's PartSetHeader is correct
@ -534,7 +534,7 @@ func (conR *ConsensusReactor) gossipDataForCatchup(logger log.Logger, rs *RoundS
}
}
func (conR *ConsensusReactor) gossipVotesRoutine(peer *p2p.Peer, ps *PeerState) {
func (conR *ConsensusReactor) gossipVotesRoutine(peer p2p.Peer, ps *PeerState) {
logger := conR.Logger.With("peer", peer)
// Simple hack to throttle logs upon sleep.
@ -644,7 +644,7 @@ func (conR *ConsensusReactor) gossipVotesForHeight(logger log.Logger, rs *RoundS
// NOTE: `queryMaj23Routine` has a simple crude design since it only comes
// into play for liveness when there's a signature DDoS attack happening.
func (conR *ConsensusReactor) queryMaj23Routine(peer *p2p.Peer, ps *PeerState) {
func (conR *ConsensusReactor) queryMaj23Routine(peer p2p.Peer, ps *PeerState) {
logger := conR.Logger.With("peer", peer)
OUTER_LOOP:
@ -743,7 +743,7 @@ func (conR *ConsensusReactor) StringIndented(indent string) string {
s := "ConsensusReactor{\n"
s += indent + " " + conR.conS.StringIndented(indent+" ") + "\n"
for _, peer := range conR.Switch.Peers().List() {
ps := peer.Data.Get(types.PeerStateKey).(*PeerState)
ps := peer.Get(types.PeerStateKey).(*PeerState)
s += indent + " " + ps.StringIndented(indent+" ") + "\n"
}
s += indent + "}"
@ -808,16 +808,18 @@ var (
// PeerState contains the known state of a peer, including its connection
// and threadsafe access to its PeerRoundState.
type PeerState struct {
Peer *p2p.Peer
Peer p2p.Peer
logger log.Logger
mtx sync.Mutex
PeerRoundState
}
// NewPeerState returns a new PeerState for the given Peer
func NewPeerState(peer *p2p.Peer) *PeerState {
func NewPeerState(peer p2p.Peer) *PeerState {
return &PeerState{
Peer: peer,
Peer: peer,
logger: log.NewNopLogger(),
PeerRoundState: PeerRoundState{
Round: -1,
ProposalPOLRound: -1,
@ -827,6 +829,11 @@ func NewPeerState(peer *p2p.Peer) *PeerState {
}
}
func (ps *PeerState) SetLogger(logger log.Logger) *PeerState {
ps.logger = logger
return ps
}
// GetRoundState returns an atomic snapshot of the PeerRoundState.
// There's no point in mutating it since it won't change PeerState.
func (ps *PeerState) GetRoundState() *PeerRoundState {
@ -1025,7 +1032,7 @@ func (ps *PeerState) SetHasVote(vote *types.Vote) {
}
func (ps *PeerState) setHasVote(height int, round int, type_ byte, index int) {
logger := ps.Peer.Logger.With("peerRound", ps.Round, "height", height, "round", round)
logger := ps.logger.With("peerRound", ps.Round, "height", height, "round", round)
logger.Debug("setHasVote(LastCommit)", "lastCommit", ps.LastCommit, "index", index)
// NOTE: some may be nil BitArrays -> no side effects.
@ -1163,7 +1170,7 @@ func (ps *PeerState) StringIndented(indent string) string {
%s Key %v
%s PRS %v
%s}`,
indent, ps.Peer.Key,
indent, ps.Peer.Key(),
indent, ps.PeerRoundState.StringIndented(indent+" "),
indent)
}


+ 3
- 3
mempool/reactor.go View File

@ -60,18 +60,18 @@ func (memR *MempoolReactor) GetChannels() []*p2p.ChannelDescriptor {
// AddPeer implements Reactor.
// It starts a broadcast routine ensuring all txs are forwarded to the given peer.
func (memR *MempoolReactor) AddPeer(peer *p2p.Peer) {
func (memR *MempoolReactor) AddPeer(peer p2p.Peer) {
go memR.broadcastTxRoutine(peer)
}
// RemovePeer implements Reactor.
func (memR *MempoolReactor) RemovePeer(peer *p2p.Peer, reason interface{}) {
func (memR *MempoolReactor) RemovePeer(peer p2p.Peer, reason interface{}) {
// broadcast routine checks if peer is gone and returns
}
// Receive implements Reactor.
// It adds any received transactions to the mempool.
func (memR *MempoolReactor) Receive(chID byte, src *p2p.Peer, msgBytes []byte) {
func (memR *MempoolReactor) Receive(chID byte, src p2p.Peer, msgBytes []byte) {
_, msg, err := DecodeMessage(msgBytes)
if err != nil {
memR.Logger.Error("Error decoding message", "err", err)


+ 71
- 33
p2p/peer.go View File

@ -12,12 +12,29 @@ import (
cmn "github.com/tendermint/tmlibs/common"
)
// Peer is an interface representing a peer connected on a reactor.
type Peer interface {
cmn.Service
Key() string
IsOutbound() bool
IsPersistent() bool
NodeInfo() *NodeInfo
Status() ConnectionStatus
Send(byte, interface{}) bool
TrySend(byte, interface{}) bool
Set(string, interface{})
Get(string) interface{}
}
// Peer could be marked as persistent, in which case you can use
// Redial function to reconnect. Note that inbound peers can't be
// made persistent. They should be made persistent on the other end.
//
// Before using a peer, you will need to perform a handshake on connection.
type Peer struct {
type peer struct {
cmn.BaseService
outbound bool
@ -28,9 +45,9 @@ type Peer struct {
persistent bool
config *PeerConfig
*NodeInfo
Key string
Data *cmn.CMap // User data.
nodeInfo *NodeInfo
key string
Data *cmn.CMap // User data.
}
// PeerConfig is a Peer configuration.
@ -60,7 +77,7 @@ func DefaultPeerConfig() *PeerConfig {
}
func newOutboundPeer(addr *NetAddress, reactorsByCh map[byte]Reactor, chDescs []*ChannelDescriptor,
onPeerError func(*Peer, interface{}), ourNodePrivKey crypto.PrivKeyEd25519, config *PeerConfig) (*Peer, error) {
onPeerError func(Peer, interface{}), ourNodePrivKey crypto.PrivKeyEd25519, config *PeerConfig) (*peer, error) {
conn, err := dial(addr, config)
if err != nil {
@ -76,13 +93,13 @@ func newOutboundPeer(addr *NetAddress, reactorsByCh map[byte]Reactor, chDescs []
}
func newInboundPeer(conn net.Conn, reactorsByCh map[byte]Reactor, chDescs []*ChannelDescriptor,
onPeerError func(*Peer, interface{}), ourNodePrivKey crypto.PrivKeyEd25519, config *PeerConfig) (*Peer, error) {
onPeerError func(Peer, interface{}), ourNodePrivKey crypto.PrivKeyEd25519, config *PeerConfig) (*peer, error) {
return newPeerFromConnAndConfig(conn, false, reactorsByCh, chDescs, onPeerError, ourNodePrivKey, config)
}
func newPeerFromConnAndConfig(rawConn net.Conn, outbound bool, reactorsByCh map[byte]Reactor, chDescs []*ChannelDescriptor,
onPeerError func(*Peer, interface{}), ourNodePrivKey crypto.PrivKeyEd25519, config *PeerConfig) (*Peer, error) {
onPeerError func(Peer, interface{}), ourNodePrivKey crypto.PrivKeyEd25519, config *PeerConfig) (*peer, error) {
conn := rawConn
@ -104,7 +121,7 @@ func newPeerFromConnAndConfig(rawConn net.Conn, outbound bool, reactorsByCh map[
}
// Key and NodeInfo are set after Handshake
p := &Peer{
p := &peer{
outbound: outbound,
conn: conn,
config: config,
@ -119,12 +136,12 @@ func newPeerFromConnAndConfig(rawConn net.Conn, outbound bool, reactorsByCh map[
}
// CloseConn should be used when the peer was created, but never started.
func (p *Peer) CloseConn() {
func (p *peer) CloseConn() {
p.conn.Close()
}
// makePersistent marks the peer as persistent.
func (p *Peer) makePersistent() {
func (p *peer) makePersistent() {
if !p.outbound {
panic("inbound peers can't be made persistent")
}
@ -133,13 +150,13 @@ func (p *Peer) makePersistent() {
}
// IsPersistent returns true if the peer is persitent, false otherwise.
func (p *Peer) IsPersistent() bool {
func (p *peer) IsPersistent() bool {
return p.persistent
}
// HandshakeTimeout performs a handshake between a given node and the peer.
// NOTE: blocking
func (p *Peer) HandshakeTimeout(ourNodeInfo *NodeInfo, timeout time.Duration) error {
func (p *peer) HandshakeTimeout(ourNodeInfo *NodeInfo, timeout time.Duration) error {
// Set deadline for handshake so we don't block forever on conn.ReadFull
p.conn.SetDeadline(time.Now().Add(timeout))
@ -176,19 +193,19 @@ func (p *Peer) HandshakeTimeout(ourNodeInfo *NodeInfo, timeout time.Duration) er
peerNodeInfo.RemoteAddr = p.Addr().String()
p.NodeInfo = peerNodeInfo
p.Key = peerNodeInfo.PubKey.KeyString()
p.nodeInfo = peerNodeInfo
p.key = peerNodeInfo.PubKey.KeyString()
return nil
}
// Addr returns peer's remote network address.
func (p *Peer) Addr() net.Addr {
func (p *peer) Addr() net.Addr {
return p.conn.RemoteAddr()
}
// PubKey returns peer's public key.
func (p *Peer) PubKey() crypto.PubKeyEd25519 {
func (p *peer) PubKey() crypto.PubKeyEd25519 {
if p.config.AuthEnc {
return p.conn.(*SecretConnection).RemotePubKey()
}
@ -199,31 +216,31 @@ func (p *Peer) PubKey() crypto.PubKeyEd25519 {
}
// OnStart implements BaseService.
func (p *Peer) OnStart() error {
func (p *peer) OnStart() error {
p.BaseService.OnStart()
_, err := p.mconn.Start()
return err
}
// OnStop implements BaseService.
func (p *Peer) OnStop() {
func (p *peer) OnStop() {
p.BaseService.OnStop()
p.mconn.Stop()
}
// Connection returns underlying MConnection.
func (p *Peer) Connection() *MConnection {
func (p *peer) Connection() *MConnection {
return p.mconn
}
// IsOutbound returns true if the connection is outbound, false otherwise.
func (p *Peer) IsOutbound() bool {
func (p *peer) IsOutbound() bool {
return p.outbound
}
// Send msg to the channel identified by chID byte. Returns false if the send
// queue is full after timeout, specified by MConnection.
func (p *Peer) Send(chID byte, msg interface{}) bool {
func (p *peer) Send(chID byte, msg interface{}) bool {
if !p.IsRunning() {
// see Switch#Broadcast, where we fetch the list of peers and loop over
// them - while we're looping, one peer may be removed and stopped.
@ -234,7 +251,7 @@ func (p *Peer) Send(chID byte, msg interface{}) bool {
// TrySend msg to the channel identified by chID byte. Immediately returns
// false if the send queue is full.
func (p *Peer) TrySend(chID byte, msg interface{}) bool {
func (p *peer) TrySend(chID byte, msg interface{}) bool {
if !p.IsRunning() {
return false
}
@ -242,7 +259,7 @@ func (p *Peer) TrySend(chID byte, msg interface{}) bool {
}
// CanSend returns true if the send queue is not full, false otherwise.
func (p *Peer) CanSend(chID byte) bool {
func (p *peer) CanSend(chID byte) bool {
if !p.IsRunning() {
return false
}
@ -250,32 +267,53 @@ func (p *Peer) CanSend(chID byte) bool {
}
// WriteTo writes the peer's public key to w.
func (p *Peer) WriteTo(w io.Writer) (n int64, err error) {
func (p *peer) WriteTo(w io.Writer) (n int64, err error) {
var n_ int
wire.WriteString(p.Key, w, &n_, &err)
wire.WriteString(p.key, w, &n_, &err)
n += int64(n_)
return
}
// String representation.
func (p *Peer) String() string {
func (p *peer) String() string {
if p.outbound {
return fmt.Sprintf("Peer{%v %v out}", p.mconn, p.Key[:12])
return fmt.Sprintf("Peer{%v %v out}", p.mconn, p.key[:12])
}
return fmt.Sprintf("Peer{%v %v in}", p.mconn, p.Key[:12])
return fmt.Sprintf("Peer{%v %v in}", p.mconn, p.key[:12])
}
// Equals reports whenever 2 peers are actually represent the same node.
func (p *Peer) Equals(other *Peer) bool {
return p.Key == other.Key
func (p *peer) Equals(other Peer) bool {
return p.key == other.Key()
}
// Get the data for a given key.
func (p *Peer) Get(key string) interface{} {
func (p *peer) Get(key string) interface{} {
return p.Data.Get(key)
}
// Set sets the data for the given key.
func (p *peer) Set(key string, data interface{}) {
p.Data.Set(key, data)
}
// Key returns the peer's id key.
func (p *peer) Key() string {
return p.key
}
// NodeInfo returns a copy of the peer's NodeInfo.
func (p *peer) NodeInfo() *NodeInfo {
n := *p.nodeInfo // copy
return &n
}
// Status returns the peer's ConnectionStatus.
func (p *peer) Status() ConnectionStatus {
return p.mconn.Status()
}
func dial(addr *NetAddress, config *PeerConfig) (net.Conn, error) {
conn, err := addr.DialTimeout(config.DialTimeout * time.Second)
if err != nil {
@ -284,8 +322,8 @@ func dial(addr *NetAddress, config *PeerConfig) (net.Conn, error) {
return conn, nil
}
func createMConnection(conn net.Conn, p *Peer, reactorsByCh map[byte]Reactor, chDescs []*ChannelDescriptor,
onPeerError func(*Peer, interface{}), config *MConnConfig) *MConnection {
func createMConnection(conn net.Conn, p *peer, reactorsByCh map[byte]Reactor, chDescs []*ChannelDescriptor,
onPeerError func(Peer, interface{}), config *MConnConfig) *MConnection {
onReceive := func(chID byte, msgBytes []byte) {
reactor := reactorsByCh[chID]


+ 16
- 16
p2p/peer_set.go View File

@ -7,8 +7,8 @@ import (
// IPeerSet has a (immutable) subset of the methods of PeerSet.
type IPeerSet interface {
Has(key string) bool
Get(key string) *Peer
List() []*Peer
Get(key string) Peer
List() []Peer
Size() int
}
@ -19,11 +19,11 @@ type IPeerSet interface {
type PeerSet struct {
mtx sync.Mutex
lookup map[string]*peerSetItem
list []*Peer
list []Peer
}
type peerSetItem struct {
peer *Peer
peer Peer
index int
}
@ -31,16 +31,16 @@ type peerSetItem struct {
func NewPeerSet() *PeerSet {
return &PeerSet{
lookup: make(map[string]*peerSetItem),
list: make([]*Peer, 0, 256),
list: make([]Peer, 0, 256),
}
}
// Add adds the peer to the PeerSet.
// It returns ErrSwitchDuplicatePeer if the peer is already present.
func (ps *PeerSet) Add(peer *Peer) error {
func (ps *PeerSet) Add(peer Peer) error {
ps.mtx.Lock()
defer ps.mtx.Unlock()
if ps.lookup[peer.Key] != nil {
if ps.lookup[peer.Key()] != nil {
return ErrSwitchDuplicatePeer
}
@ -48,7 +48,7 @@ func (ps *PeerSet) Add(peer *Peer) error {
// Appending is safe even with other goroutines
// iterating over the ps.list slice.
ps.list = append(ps.list, peer)
ps.lookup[peer.Key] = &peerSetItem{peer, index}
ps.lookup[peer.Key()] = &peerSetItem{peer, index}
return nil
}
@ -62,7 +62,7 @@ func (ps *PeerSet) Has(peerKey string) bool {
}
// Get looks up a peer by the provided peerKey.
func (ps *PeerSet) Get(peerKey string) *Peer {
func (ps *PeerSet) Get(peerKey string) Peer {
ps.mtx.Lock()
defer ps.mtx.Unlock()
item, ok := ps.lookup[peerKey]
@ -74,10 +74,10 @@ func (ps *PeerSet) Get(peerKey string) *Peer {
}
// Remove discards peer by its Key, if the peer was previously memoized.
func (ps *PeerSet) Remove(peer *Peer) {
func (ps *PeerSet) Remove(peer Peer) {
ps.mtx.Lock()
defer ps.mtx.Unlock()
item := ps.lookup[peer.Key]
item := ps.lookup[peer.Key()]
if item == nil {
return
}
@ -85,23 +85,23 @@ func (ps *PeerSet) Remove(peer *Peer) {
index := item.index
// Create a new copy of the list but with one less item.
// (we must copy because we'll be mutating the list).
newList := make([]*Peer, len(ps.list)-1)
newList := make([]Peer, len(ps.list)-1)
copy(newList, ps.list)
// If it's the last peer, that's an easy special case.
if index == len(ps.list)-1 {
ps.list = newList
delete(ps.lookup, peer.Key)
delete(ps.lookup, peer.Key())
return
}
// Replace the popped item with the last item in the old list.
lastPeer := ps.list[len(ps.list)-1]
lastPeerKey := lastPeer.Key
lastPeerKey := lastPeer.Key()
lastPeerItem := ps.lookup[lastPeerKey]
newList[index] = lastPeer
lastPeerItem.index = index
ps.list = newList
delete(ps.lookup, peer.Key)
delete(ps.lookup, peer.Key())
}
// Size returns the number of unique items in the peerSet.
@ -112,7 +112,7 @@ func (ps *PeerSet) Size() int {
}
// List returns the threadsafe list of peers.
func (ps *PeerSet) List() []*Peer {
func (ps *PeerSet) List() []Peer {
ps.mtx.Lock()
defer ps.mtx.Unlock()
return ps.list


+ 11
- 11
p2p/peer_set_test.go View File

@ -11,10 +11,10 @@ import (
)
// Returns an empty dummy peer
func randPeer() *Peer {
return &Peer{
Key: cmn.RandStr(12),
NodeInfo: &NodeInfo{
func randPeer() *peer {
return &peer{
key: cmn.RandStr(12),
nodeInfo: &NodeInfo{
RemoteAddr: cmn.Fmt("%v.%v.%v.%v:46656", rand.Int()%256, rand.Int()%256, rand.Int()%256, rand.Int()%256),
ListenAddr: cmn.Fmt("%v.%v.%v.%v:46656", rand.Int()%256, rand.Int()%256, rand.Int()%256, rand.Int()%256),
},
@ -25,7 +25,7 @@ func TestPeerSetAddRemoveOne(t *testing.T) {
t.Parallel()
peerSet := NewPeerSet()
var peerList []*Peer
var peerList []Peer
for i := 0; i < 5; i++ {
p := randPeer()
peerSet.Add(p)
@ -38,7 +38,7 @@ func TestPeerSetAddRemoveOne(t *testing.T) {
peerSet.Remove(peerAtFront)
wantSize := n - i - 1
for j := 0; j < 2; j++ {
assert.Equal(t, false, peerSet.Has(peerAtFront.Key), "#%d Run #%d: failed to remove peer", i, j)
assert.Equal(t, false, peerSet.Has(peerAtFront.Key()), "#%d Run #%d: failed to remove peer", i, j)
assert.Equal(t, wantSize, peerSet.Size(), "#%d Run #%d: failed to remove peer and decrement size", i, j)
// Test the route of removing the now non-existent element
peerSet.Remove(peerAtFront)
@ -55,7 +55,7 @@ func TestPeerSetAddRemoveOne(t *testing.T) {
for i := n - 1; i >= 0; i-- {
peerAtEnd := peerList[i]
peerSet.Remove(peerAtEnd)
assert.Equal(t, false, peerSet.Has(peerAtEnd.Key), "#%d: failed to remove item at end", i)
assert.Equal(t, false, peerSet.Has(peerAtEnd.Key()), "#%d: failed to remove item at end", i)
assert.Equal(t, i, peerSet.Size(), "#%d: differing sizes after peerSet.Remove(atEndPeer)", i)
}
}
@ -64,7 +64,7 @@ func TestPeerSetAddRemoveMany(t *testing.T) {
t.Parallel()
peerSet := NewPeerSet()
peers := []*Peer{}
peers := []Peer{}
N := 100
for i := 0; i < N; i++ {
peer := randPeer()
@ -79,7 +79,7 @@ func TestPeerSetAddRemoveMany(t *testing.T) {
for i, peer := range peers {
peerSet.Remove(peer)
if peerSet.Has(peer.Key) {
if peerSet.Has(peer.Key()) {
t.Errorf("Failed to remove peer")
}
if peerSet.Size() != len(peers)-i-1 {
@ -126,7 +126,7 @@ func TestPeerSetGet(t *testing.T) {
t.Parallel()
peerSet := NewPeerSet()
peer := randPeer()
assert.Nil(t, peerSet.Get(peer.Key), "expecting a nil lookup, before .Add")
assert.Nil(t, peerSet.Get(peer.Key()), "expecting a nil lookup, before .Add")
if err := peerSet.Add(peer); err != nil {
t.Fatalf("Failed to add new peer: %v", err)
@ -139,7 +139,7 @@ func TestPeerSetGet(t *testing.T) {
wg.Add(1)
go func(i int) {
defer wg.Done()
got, want := peerSet.Get(peer.Key), peer
got, want := peerSet.Get(peer.Key()), peer
assert.Equal(t, got, want, "#%d: got=%v want=%v", i, got, want)
}(i)
}


+ 3
- 3
p2p/peer_test.go View File

@ -76,13 +76,13 @@ func TestPeerSend(t *testing.T) {
assert.True(p.Send(0x01, "Asylum"))
}
func createOutboundPeerAndPerformHandshake(addr *NetAddress, config *PeerConfig) (*Peer, error) {
func createOutboundPeerAndPerformHandshake(addr *NetAddress, config *PeerConfig) (*peer, error) {
chDescs := []*ChannelDescriptor{
&ChannelDescriptor{ID: 0x01, Priority: 1},
}
reactorsByCh := map[byte]Reactor{0x01: NewTestReactor(chDescs, true)}
pk := crypto.GenPrivKeyEd25519()
p, err := newOutboundPeer(addr, reactorsByCh, chDescs, func(p *Peer, r interface{}) {}, pk, config)
p, err := newOutboundPeer(addr, reactorsByCh, chDescs, func(p Peer, r interface{}) {}, pk, config)
if err != nil {
return nil, err
}
@ -133,7 +133,7 @@ func (p *remotePeer) accept(l net.Listener) {
if err != nil {
golog.Fatalf("Failed to accept conn: %+v", err)
}
peer, err := newInboundPeer(conn, make(map[byte]Reactor), make([]*ChannelDescriptor, 0), func(p *Peer, r interface{}) {}, p.PrivKey, p.Config)
peer, err := newInboundPeer(conn, make(map[byte]Reactor), make([]*ChannelDescriptor, 0), func(p Peer, r interface{}) {}, p.PrivKey, p.Config)
if err != nil {
golog.Fatalf("Failed to create a peer: %+v", err)
}


+ 9
- 9
p2p/pex_reactor.go View File

@ -92,7 +92,7 @@ func (r *PEXReactor) GetChannels() []*ChannelDescriptor {
// AddPeer implements Reactor by adding peer to the address book (if inbound)
// or by requesting more addresses (if outbound).
func (r *PEXReactor) AddPeer(p *Peer) {
func (r *PEXReactor) AddPeer(p Peer) {
if p.IsOutbound() {
// For outbound peers, the address is already in the books.
// Either it was added in DialSeeds or when we
@ -101,10 +101,10 @@ func (r *PEXReactor) AddPeer(p *Peer) {
r.RequestPEX(p)
}
} else { // For inbound connections, the peer is its own source
addr, err := NewNetAddressString(p.ListenAddr)
addr, err := NewNetAddressString(p.NodeInfo().ListenAddr)
if err != nil {
// this should never happen
r.Logger.Error("Error in AddPeer: invalid peer address", "addr", p.ListenAddr, "err", err)
r.Logger.Error("Error in AddPeer: invalid peer address", "addr", p.NodeInfo().ListenAddr, "err", err)
return
}
r.book.AddAddress(addr, addr)
@ -112,15 +112,15 @@ func (r *PEXReactor) AddPeer(p *Peer) {
}
// RemovePeer implements Reactor.
func (r *PEXReactor) RemovePeer(p *Peer, reason interface{}) {
func (r *PEXReactor) RemovePeer(p Peer, reason interface{}) {
// If we aren't keeping track of local temp data for each peer here, then we
// don't have to do anything.
}
// Receive implements Reactor by handling incoming PEX messages.
func (r *PEXReactor) Receive(chID byte, src *Peer, msgBytes []byte) {
srcAddr := src.Connection().RemoteAddress
srcAddrStr := srcAddr.String()
func (r *PEXReactor) Receive(chID byte, src Peer, msgBytes []byte) {
srcAddrStr := src.NodeInfo().RemoteAddr
srcAddr, _ := NewNetAddressString(srcAddrStr)
r.IncrementMsgCountForPeer(srcAddrStr)
if r.ReachedMaxMsgCountForPeer(srcAddrStr) {
@ -154,12 +154,12 @@ func (r *PEXReactor) Receive(chID byte, src *Peer, msgBytes []byte) {
}
// RequestPEX asks peer for more addresses.
func (r *PEXReactor) RequestPEX(p *Peer) {
func (r *PEXReactor) RequestPEX(p Peer) {
p.Send(PexChannel, struct{ PexMessage }{&pexRequestMessage{}})
}
// SendAddrs sends addrs to the peer.
func (r *PEXReactor) SendAddrs(p *Peer, addrs []*NetAddress) {
func (r *PEXReactor) SendAddrs(p Peer, addrs []*NetAddress) {
p.Send(PexChannel, struct{ PexMessage }{&pexAddrsMessage{Addrs: addrs}})
}


+ 7
- 6
p2p/pex_reactor_test.go View File

@ -129,7 +129,7 @@ func TestPEXReactorReceive(t *testing.T) {
peer := createRandomPeer(false)
size := book.Size()
netAddr, _ := NewNetAddressString(peer.ListenAddr)
netAddr, _ := NewNetAddressString(peer.NodeInfo().ListenAddr)
addrs := []*NetAddress{netAddr}
msg := wire.BinaryBytes(struct{ PexMessage }{&pexAddrsMessage{Addrs: addrs}})
r.Receive(PexChannel, peer, msg)
@ -159,16 +159,17 @@ func TestPEXReactorAbuseFromPeer(t *testing.T) {
r.Receive(PexChannel, peer, msg)
}
assert.True(r.ReachedMaxMsgCountForPeer(peer.ListenAddr))
assert.True(r.ReachedMaxMsgCountForPeer(peer.NodeInfo().ListenAddr))
}
func createRandomPeer(outbound bool) *Peer {
func createRandomPeer(outbound bool) *peer {
addr := cmn.Fmt("%v.%v.%v.%v:46656", rand.Int()%256, rand.Int()%256, rand.Int()%256, rand.Int()%256)
netAddr, _ := NewNetAddressString(addr)
p := &Peer{
Key: cmn.RandStr(12),
NodeInfo: &NodeInfo{
p := &peer{
key: cmn.RandStr(12),
nodeInfo: &NodeInfo{
ListenAddr: addr,
RemoteAddr: netAddr.String(),
},
outbound: outbound,
mconn: &MConnection{RemoteAddress: netAddr},


+ 21
- 20
p2p/switch.go View File

@ -22,9 +22,9 @@ type Reactor interface {
SetSwitch(*Switch)
GetChannels() []*ChannelDescriptor
AddPeer(peer *Peer)
RemovePeer(peer *Peer, reason interface{})
Receive(chID byte, peer *Peer, msgBytes []byte)
AddPeer(peer Peer)
RemovePeer(peer Peer, reason interface{})
Receive(chID byte, peer Peer, msgBytes []byte)
}
//--------------------------------------
@ -44,10 +44,10 @@ func NewBaseReactor(name string, impl Reactor) *BaseReactor {
func (br *BaseReactor) SetSwitch(sw *Switch) {
br.Switch = sw
}
func (_ *BaseReactor) GetChannels() []*ChannelDescriptor { return nil }
func (_ *BaseReactor) AddPeer(peer *Peer) {}
func (_ *BaseReactor) RemovePeer(peer *Peer, reason interface{}) {}
func (_ *BaseReactor) Receive(chID byte, peer *Peer, msgBytes []byte) {}
func (_ *BaseReactor) GetChannels() []*ChannelDescriptor { return nil }
func (_ *BaseReactor) AddPeer(peer Peer) {}
func (_ *BaseReactor) RemovePeer(peer Peer, reason interface{}) {}
func (_ *BaseReactor) Receive(chID byte, peer Peer, msgBytes []byte) {}
//-----------------------------------------------------------------------------
@ -213,7 +213,8 @@ func (sw *Switch) OnStop() {
// and to all registered reactors.
// NOTE: This performs a blocking handshake before the peer is added.
// CONTRACT: If error is returned, peer is nil, and conn is immediately closed.
func (sw *Switch) AddPeer(peer *Peer) error {
func (sw *Switch) AddPeer(peer *peer) error {
if err := sw.FilterConnByAddr(peer.Addr()); err != nil {
return err
}
@ -232,12 +233,12 @@ func (sw *Switch) AddPeer(peer *Peer) error {
}
// Check version, chain id
if err := sw.nodeInfo.CompatibleWith(peer.NodeInfo); err != nil {
if err := sw.nodeInfo.CompatibleWith(peer.NodeInfo()); err != nil {
return err
}
// Check for duplicate peer
if sw.peers.Has(peer.Key) {
if sw.peers.Has(peer.Key()) {
return ErrSwitchDuplicatePeer
}
@ -285,7 +286,7 @@ func (sw *Switch) SetPubKeyFilter(f func(crypto.PubKeyEd25519) error) {
sw.filterConnByPubKey = f
}
func (sw *Switch) startInitPeer(peer *Peer) {
func (sw *Switch) startInitPeer(peer *peer) {
peer.Start() // spawn send/recv routines
for _, reactor := range sw.reactors {
reactor.AddPeer(peer)
@ -337,7 +338,7 @@ func (sw *Switch) dialSeed(addr *NetAddress) {
// DialPeerWithAddress dials the given peer and runs sw.AddPeer if it connects successfully.
// If `persistent == true`, the switch will always try to reconnect to this peer if the connection ever fails.
func (sw *Switch) DialPeerWithAddress(addr *NetAddress, persistent bool) (*Peer, error) {
func (sw *Switch) DialPeerWithAddress(addr *NetAddress, persistent bool) (Peer, error) {
sw.dialing.Set(addr.IP.String(), addr)
defer sw.dialing.Delete(addr.IP.String())
@ -375,7 +376,7 @@ func (sw *Switch) Broadcast(chID byte, msg interface{}) chan bool {
successChan := make(chan bool, len(sw.peers.List()))
sw.Logger.Debug("Broadcast", "channel", chID, "msg", msg)
for _, peer := range sw.peers.List() {
go func(peer *Peer) {
go func(peer Peer) {
success := peer.Send(chID, msg)
successChan <- success
}(peer)
@ -387,7 +388,7 @@ func (sw *Switch) Broadcast(chID byte, msg interface{}) chan bool {
func (sw *Switch) NumPeers() (outbound, inbound, dialing int) {
peers := sw.peers.List()
for _, peer := range peers {
if peer.outbound {
if peer.IsOutbound() {
outbound++
} else {
inbound++
@ -405,8 +406,8 @@ func (sw *Switch) Peers() IPeerSet {
// StopPeerForError disconnects from a peer due to external error.
// If the peer is persistent, it will attempt to reconnect.
// TODO: make record depending on reason.
func (sw *Switch) StopPeerForError(peer *Peer, reason interface{}) {
addr := NewNetAddress(peer.Addr())
func (sw *Switch) StopPeerForError(peer Peer, reason interface{}) {
addr, _ := NewNetAddressString(peer.NodeInfo().RemoteAddr)
sw.Logger.Error("Stopping peer for error", "peer", peer, "err", reason)
sw.stopAndRemovePeer(peer, reason)
@ -438,12 +439,12 @@ func (sw *Switch) StopPeerForError(peer *Peer, reason interface{}) {
// StopPeerGracefully disconnects from a peer gracefully.
// TODO: handle graceful disconnects.
func (sw *Switch) StopPeerGracefully(peer *Peer) {
func (sw *Switch) StopPeerGracefully(peer Peer) {
sw.Logger.Info("Stopping peer gracefully")
sw.stopAndRemovePeer(peer, nil)
}
func (sw *Switch) stopAndRemovePeer(peer *Peer, reason interface{}) {
func (sw *Switch) stopAndRemovePeer(peer Peer, reason interface{}) {
sw.peers.Remove(peer)
peer.Stop()
for _, reactor := range sw.reactors {
@ -483,11 +484,11 @@ func (sw *Switch) listenerRoutine(l Listener) {
//-----------------------------------------------------------------------------
type SwitchEventNewPeer struct {
Peer *Peer
Peer Peer
}
type SwitchEventDonePeer struct {
Peer *Peer
Peer Peer
Error interface{}
}


+ 6
- 6
p2p/switch_test.go View File

@ -37,8 +37,8 @@ type TestReactor struct {
mtx sync.Mutex
channels []*ChannelDescriptor
peersAdded []*Peer
peersRemoved []*Peer
peersAdded []Peer
peersRemoved []Peer
logMessages bool
msgsCounter int
msgsReceived map[byte][]PeerMessage
@ -59,24 +59,24 @@ func (tr *TestReactor) GetChannels() []*ChannelDescriptor {
return tr.channels
}
func (tr *TestReactor) AddPeer(peer *Peer) {
func (tr *TestReactor) AddPeer(peer Peer) {
tr.mtx.Lock()
defer tr.mtx.Unlock()
tr.peersAdded = append(tr.peersAdded, peer)
}
func (tr *TestReactor) RemovePeer(peer *Peer, reason interface{}) {
func (tr *TestReactor) RemovePeer(peer Peer, reason interface{}) {
tr.mtx.Lock()
defer tr.mtx.Unlock()
tr.peersRemoved = append(tr.peersRemoved, peer)
}
func (tr *TestReactor) Receive(chID byte, peer *Peer, msgBytes []byte) {
func (tr *TestReactor) Receive(chID byte, peer Peer, msgBytes []byte) {
if tr.logMessages {
tr.mtx.Lock()
defer tr.mtx.Unlock()
//fmt.Printf("Received: %X, %X\n", chID, msgBytes)
tr.msgsReceived[chID] = append(tr.msgsReceived[chID], PeerMessage{peer.Key, msgBytes, tr.msgsCounter})
tr.msgsReceived[chID] = append(tr.msgsReceived[chID], PeerMessage{peer.Key(), msgBytes, tr.msgsCounter})
tr.msgsCounter++
}
}


+ 2
- 2
rpc/core/consensus.go View File

@ -86,9 +86,9 @@ func DumpConsensusState() (*ctypes.ResultDumpConsensusState, error) {
peerRoundStates := []string{}
for _, peer := range p2pSwitch.Peers().List() {
// TODO: clean this up?
peerState := peer.Data.Get(types.PeerStateKey).(*cm.PeerState)
peerState := peer.Get(types.PeerStateKey).(*cm.PeerState)
peerRoundState := peerState.GetRoundState()
peerRoundStateStr := peer.Key + ":" + string(wire.JSONBytes(peerRoundState))
peerRoundStateStr := peer.Key() + ":" + string(wire.JSONBytes(peerRoundState))
peerRoundStates = append(peerRoundStates, peerRoundStateStr)
}
return &ctypes.ResultDumpConsensusState{roundState.String(), peerRoundStates}, nil


+ 2
- 2
rpc/core/net.go View File

@ -42,9 +42,9 @@ func NetInfo() (*ctypes.ResultNetInfo, error) {
peers := []ctypes.Peer{}
for _, peer := range p2pSwitch.Peers().List() {
peers = append(peers, ctypes.Peer{
NodeInfo: *peer.NodeInfo,
NodeInfo: *peer.NodeInfo(),
IsOutbound: peer.IsOutbound(),
ConnectionStatus: peer.Connection().Status(),
ConnectionStatus: peer.Status(),
})
}
return &ctypes.ResultNetInfo{


Loading…
Cancel
Save