mirror of
https://gitlab.com/pulsechaincom/prysm-pulse.git
synced 2024-12-24 12:27:18 +00:00
cec32cb996
* Append Dynamic Addinng Trusted Peer Apis * Append unit tests for Dynamic Addinng Trusted Peer Apis * Update beacon-chain/p2p/peers/peerdata/store.go * Update beacon-chain/p2p/peers/peerdata/store_test.go * Update beacon-chain/p2p/peers/peerdata/store_test.go * Update beacon-chain/p2p/peers/peerdata/store_test.go * Update beacon-chain/p2p/peers/status.go * Update beacon-chain/p2p/peers/status_test.go * Update beacon-chain/p2p/peers/status_test.go * Update beacon-chain/rpc/eth/node/handlers.go * Update beacon-chain/rpc/eth/node/handlers.go * Update beacon-chain/rpc/eth/node/handlers.go * Update beacon-chain/rpc/eth/node/handlers.go * Move trusted peer apis from rpc/eth/v1/node to rpc/prysm/node; move peersToWatch into ensurePeerConnections function; * Update beacon-chain/rpc/prysm/node/server.go * Update beacon-chain/rpc/prysm/node/server.go * fix go lint problem * p2p/watch_peers.go: trusted peer makes AddrInfo structure by itself instead of using MakePeer(). p2p/service.go: add connectWithAllTrustedPeers function, before connectWithPeer, add trusted peer info into peer status. p2p/peers/status.go: trusted peers are not included, when pruning outdated and disconnected peers. * use readlock for GetTrustedPeers and IsTrustedPeers --------- Co-authored-by: simon <sunminghui2@huawei.com> Co-authored-by: Radosław Kapka <rkapka@wp.pl> Co-authored-by: Nishant Das <nishdas93@gmail.com>
1039 lines
31 KiB
Go
1039 lines
31 KiB
Go
// Package peers provides information about peers at the Ethereum consensus protocol level.
|
|
//
|
|
// "Protocol level" is the level above the network level, so this layer never sees or interacts with
|
|
// (for example) hosts that are uncontactable due to being down, firewalled, etc. Instead, this works
|
|
// with peers that are contactable but may or may not be of the correct fork version, not currently
|
|
// required due to the number of current connections, etc.
|
|
//
|
|
// A peer can have one of a number of states:
|
|
//
|
|
// - connected if we are able to talk to the remote peer
|
|
// - connecting if we are attempting to be able to talk to the remote peer
|
|
// - disconnecting if we are attempting to stop being able to talk to the remote peer
|
|
// - disconnected if we are not able to talk to the remote peer
|
|
//
|
|
// For convenience, there are two aggregate states expressed in functions:
|
|
//
|
|
// - active if we are connecting or connected
|
|
// - inactive if we are disconnecting or disconnected
|
|
//
|
|
// Peer information is persistent for the run of the service. This allows for collection of useful
|
|
// long-term statistics such as number of bad responses obtained from the peer, giving the basis for
|
|
// decisions to not talk to known-bad peers (by de-scoring them).
|
|
package peers
|
|
|
|
import (
|
|
"context"
|
|
"math"
|
|
"sort"
|
|
"time"
|
|
|
|
"github.com/ethereum/go-ethereum/p2p/enr"
|
|
"github.com/libp2p/go-libp2p/core/network"
|
|
"github.com/libp2p/go-libp2p/core/peer"
|
|
ma "github.com/multiformats/go-multiaddr"
|
|
manet "github.com/multiformats/go-multiaddr/net"
|
|
"github.com/prysmaticlabs/go-bitfield"
|
|
"github.com/prysmaticlabs/prysm/v4/beacon-chain/p2p/peers/peerdata"
|
|
"github.com/prysmaticlabs/prysm/v4/beacon-chain/p2p/peers/scorers"
|
|
"github.com/prysmaticlabs/prysm/v4/config/features"
|
|
"github.com/prysmaticlabs/prysm/v4/config/params"
|
|
"github.com/prysmaticlabs/prysm/v4/consensus-types/primitives"
|
|
"github.com/prysmaticlabs/prysm/v4/crypto/rand"
|
|
pmath "github.com/prysmaticlabs/prysm/v4/math"
|
|
pb "github.com/prysmaticlabs/prysm/v4/proto/prysm/v1alpha1"
|
|
"github.com/prysmaticlabs/prysm/v4/proto/prysm/v1alpha1/metadata"
|
|
prysmTime "github.com/prysmaticlabs/prysm/v4/time"
|
|
"github.com/prysmaticlabs/prysm/v4/time/slots"
|
|
)
|
|
|
|
const (
|
|
// PeerDisconnected means there is no connection to the peer.
|
|
PeerDisconnected peerdata.PeerConnectionState = iota
|
|
// PeerDisconnecting means there is an on-going attempt to disconnect from the peer.
|
|
PeerDisconnecting
|
|
// PeerConnected means the peer has an active connection.
|
|
PeerConnected
|
|
// PeerConnecting means there is an on-going attempt to connect to the peer.
|
|
PeerConnecting
|
|
)
|
|
|
|
const (
|
|
// ColocationLimit restricts how many peer identities we can see from a single ip or ipv6 subnet.
|
|
ColocationLimit = 5
|
|
|
|
// Additional buffer beyond current peer limit, from which we can store the relevant peer statuses.
|
|
maxLimitBuffer = 150
|
|
|
|
// InboundRatio is the proportion of our connected peer limit at which we will allow inbound peers.
|
|
InboundRatio = float64(0.8)
|
|
|
|
// MinBackOffDuration minimum amount (in milliseconds) to wait before peer is re-dialed.
|
|
// When node and peer are dialing each other simultaneously connection may fail. In order, to break
|
|
// of constant dialing, peer is assigned some backoff period, and only dialed again once that backoff is up.
|
|
MinBackOffDuration = 100
|
|
// MaxBackOffDuration maximum amount (in milliseconds) to wait before peer is re-dialed.
|
|
MaxBackOffDuration = 5000
|
|
)
|
|
|
|
// Status is the structure holding the peer status information.
|
|
type Status struct {
|
|
ctx context.Context
|
|
scorers *scorers.Service
|
|
store *peerdata.Store
|
|
ipTracker map[string]uint64
|
|
rand *rand.Rand
|
|
}
|
|
|
|
// StatusConfig represents peer status service params.
|
|
type StatusConfig struct {
|
|
// PeerLimit specifies maximum amount of concurrent peers that are expected to be connect to the node.
|
|
PeerLimit int
|
|
// ScorerParams holds peer scorer configuration params.
|
|
ScorerParams *scorers.Config
|
|
}
|
|
|
|
// NewStatus creates a new status entity.
|
|
func NewStatus(ctx context.Context, config *StatusConfig) *Status {
|
|
store := peerdata.NewStore(ctx, &peerdata.StoreConfig{
|
|
MaxPeers: maxLimitBuffer + config.PeerLimit,
|
|
})
|
|
return &Status{
|
|
ctx: ctx,
|
|
store: store,
|
|
scorers: scorers.NewService(ctx, store, config.ScorerParams),
|
|
ipTracker: map[string]uint64{},
|
|
// Random generator used to calculate dial backoff period.
|
|
// It is ok to use deterministic generator, no need for true entropy.
|
|
rand: rand.NewDeterministicGenerator(),
|
|
}
|
|
}
|
|
|
|
// Scorers exposes peer scoring management service.
|
|
func (p *Status) Scorers() *scorers.Service {
|
|
return p.scorers
|
|
}
|
|
|
|
// MaxPeerLimit returns the max peer limit stored in the current peer store.
|
|
func (p *Status) MaxPeerLimit() int {
|
|
return p.store.Config().MaxPeers
|
|
}
|
|
|
|
// Add adds a peer.
|
|
// If a peer already exists with this ID its address and direction are updated with the supplied data.
|
|
func (p *Status) Add(record *enr.Record, pid peer.ID, address ma.Multiaddr, direction network.Direction) {
|
|
p.store.Lock()
|
|
defer p.store.Unlock()
|
|
|
|
if peerData, ok := p.store.PeerData(pid); ok {
|
|
// Peer already exists, just update its address info.
|
|
prevAddress := peerData.Address
|
|
peerData.Address = address
|
|
peerData.Direction = direction
|
|
if record != nil {
|
|
peerData.Enr = record
|
|
}
|
|
if !sameIP(prevAddress, address) {
|
|
p.addIpToTracker(pid)
|
|
}
|
|
return
|
|
}
|
|
peerData := &peerdata.PeerData{
|
|
Address: address,
|
|
Direction: direction,
|
|
// Peers start disconnected; state will be updated when the handshake process begins.
|
|
ConnState: PeerDisconnected,
|
|
}
|
|
if record != nil {
|
|
peerData.Enr = record
|
|
}
|
|
p.store.SetPeerData(pid, peerData)
|
|
p.addIpToTracker(pid)
|
|
}
|
|
|
|
// Address returns the multiaddress of the given remote peer.
|
|
// This will error if the peer does not exist.
|
|
func (p *Status) Address(pid peer.ID) (ma.Multiaddr, error) {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
|
|
if peerData, ok := p.store.PeerData(pid); ok {
|
|
return peerData.Address, nil
|
|
}
|
|
return nil, peerdata.ErrPeerUnknown
|
|
}
|
|
|
|
// Direction returns the direction of the given remote peer.
|
|
// This will error if the peer does not exist.
|
|
func (p *Status) Direction(pid peer.ID) (network.Direction, error) {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
|
|
if peerData, ok := p.store.PeerData(pid); ok {
|
|
return peerData.Direction, nil
|
|
}
|
|
return network.DirUnknown, peerdata.ErrPeerUnknown
|
|
}
|
|
|
|
// ENR returns the enr for the corresponding peer id.
|
|
func (p *Status) ENR(pid peer.ID) (*enr.Record, error) {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
|
|
if peerData, ok := p.store.PeerData(pid); ok {
|
|
return peerData.Enr, nil
|
|
}
|
|
return nil, peerdata.ErrPeerUnknown
|
|
}
|
|
|
|
// SetChainState sets the chain state of the given remote peer.
|
|
func (p *Status) SetChainState(pid peer.ID, chainState *pb.Status) {
|
|
p.scorers.PeerStatusScorer().SetPeerStatus(pid, chainState, nil)
|
|
}
|
|
|
|
// ChainState gets the chain state of the given remote peer.
|
|
// This will error if the peer does not exist.
|
|
// This will error if there is no known chain state for the peer.
|
|
func (p *Status) ChainState(pid peer.ID) (*pb.Status, error) {
|
|
return p.scorers.PeerStatusScorer().PeerStatus(pid)
|
|
}
|
|
|
|
// IsActive checks if a peers is active and returns the result appropriately.
|
|
func (p *Status) IsActive(pid peer.ID) bool {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
|
|
peerData, ok := p.store.PeerData(pid)
|
|
return ok && (peerData.ConnState == PeerConnected || peerData.ConnState == PeerConnecting)
|
|
}
|
|
|
|
// IsAboveInboundLimit checks if we are above our current inbound
|
|
// peer limit.
|
|
func (p *Status) IsAboveInboundLimit() bool {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
totalInbound := 0
|
|
for _, peerData := range p.store.Peers() {
|
|
if peerData.ConnState == PeerConnected &&
|
|
peerData.Direction == network.DirInbound {
|
|
totalInbound += 1
|
|
}
|
|
}
|
|
inboundLimit := int(float64(p.ConnectedPeerLimit()) * InboundRatio)
|
|
return totalInbound > inboundLimit
|
|
}
|
|
|
|
// InboundLimit returns the current inbound
|
|
// peer limit.
|
|
func (p *Status) InboundLimit() int {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
return int(float64(p.ConnectedPeerLimit()) * InboundRatio)
|
|
}
|
|
|
|
// SetMetadata sets the metadata of the given remote peer.
|
|
func (p *Status) SetMetadata(pid peer.ID, metaData metadata.Metadata) {
|
|
p.store.Lock()
|
|
defer p.store.Unlock()
|
|
|
|
peerData := p.store.PeerDataGetOrCreate(pid)
|
|
peerData.MetaData = metaData.Copy()
|
|
}
|
|
|
|
// Metadata returns a copy of the metadata corresponding to the provided
|
|
// peer id.
|
|
func (p *Status) Metadata(pid peer.ID) (metadata.Metadata, error) {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
|
|
if peerData, ok := p.store.PeerData(pid); ok {
|
|
if peerData.MetaData == nil || peerData.MetaData.IsNil() {
|
|
return nil, nil
|
|
}
|
|
return peerData.MetaData.Copy(), nil
|
|
}
|
|
return nil, peerdata.ErrPeerUnknown
|
|
}
|
|
|
|
// CommitteeIndices retrieves the committee subnets the peer is subscribed to.
|
|
func (p *Status) CommitteeIndices(pid peer.ID) ([]uint64, error) {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
|
|
if peerData, ok := p.store.PeerData(pid); ok {
|
|
if peerData.Enr == nil || peerData.MetaData == nil || peerData.MetaData.IsNil() {
|
|
return []uint64{}, nil
|
|
}
|
|
return indicesFromBitfield(peerData.MetaData.AttnetsBitfield()), nil
|
|
}
|
|
return nil, peerdata.ErrPeerUnknown
|
|
}
|
|
|
|
// SubscribedToSubnet retrieves the peers subscribed to the given
|
|
// committee subnet.
|
|
func (p *Status) SubscribedToSubnet(index uint64) []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
|
|
peers := make([]peer.ID, 0)
|
|
for pid, peerData := range p.store.Peers() {
|
|
// look at active peers
|
|
connectedStatus := peerData.ConnState == PeerConnecting || peerData.ConnState == PeerConnected
|
|
if connectedStatus && peerData.MetaData != nil && !peerData.MetaData.IsNil() && peerData.MetaData.AttnetsBitfield() != nil {
|
|
indices := indicesFromBitfield(peerData.MetaData.AttnetsBitfield())
|
|
for _, idx := range indices {
|
|
if idx == index {
|
|
peers = append(peers, pid)
|
|
break
|
|
}
|
|
}
|
|
}
|
|
}
|
|
return peers
|
|
}
|
|
|
|
// SetConnectionState sets the connection state of the given remote peer.
|
|
func (p *Status) SetConnectionState(pid peer.ID, state peerdata.PeerConnectionState) {
|
|
p.store.Lock()
|
|
defer p.store.Unlock()
|
|
|
|
peerData := p.store.PeerDataGetOrCreate(pid)
|
|
peerData.ConnState = state
|
|
}
|
|
|
|
// ConnectionState gets the connection state of the given remote peer.
|
|
// This will error if the peer does not exist.
|
|
func (p *Status) ConnectionState(pid peer.ID) (peerdata.PeerConnectionState, error) {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
|
|
if peerData, ok := p.store.PeerData(pid); ok {
|
|
return peerData.ConnState, nil
|
|
}
|
|
return PeerDisconnected, peerdata.ErrPeerUnknown
|
|
}
|
|
|
|
// ChainStateLastUpdated gets the last time the chain state of the given remote peer was updated.
|
|
// This will error if the peer does not exist.
|
|
func (p *Status) ChainStateLastUpdated(pid peer.ID) (time.Time, error) {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
|
|
if peerData, ok := p.store.PeerData(pid); ok {
|
|
return peerData.ChainStateLastUpdated, nil
|
|
}
|
|
return prysmTime.Now(), peerdata.ErrPeerUnknown
|
|
}
|
|
|
|
// IsBad states if the peer is to be considered bad (by *any* of the registered scorers).
|
|
// If the peer is unknown this will return `false`, which makes using this function easier than returning an error.
|
|
func (p *Status) IsBad(pid peer.ID) bool {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
return p.isBad(pid)
|
|
}
|
|
|
|
// isBad is the lock-free version of IsBad.
|
|
func (p *Status) isBad(pid peer.ID) bool {
|
|
// Do not disconnect from trusted peers.
|
|
if p.store.IsTrustedPeer(pid) {
|
|
return false
|
|
}
|
|
return p.isfromBadIP(pid) || p.scorers.IsBadPeerNoLock(pid)
|
|
}
|
|
|
|
// NextValidTime gets the earliest possible time it is to contact/dial
|
|
// a peer again. This is used to back-off from peers in the event
|
|
// they are 'full' or have banned us.
|
|
func (p *Status) NextValidTime(pid peer.ID) (time.Time, error) {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
|
|
if peerData, ok := p.store.PeerData(pid); ok {
|
|
return peerData.NextValidTime, nil
|
|
}
|
|
return prysmTime.Now(), peerdata.ErrPeerUnknown
|
|
}
|
|
|
|
// SetNextValidTime sets the earliest possible time we are
|
|
// able to contact this peer again.
|
|
func (p *Status) SetNextValidTime(pid peer.ID, nextTime time.Time) {
|
|
p.store.Lock()
|
|
defer p.store.Unlock()
|
|
|
|
peerData := p.store.PeerDataGetOrCreate(pid)
|
|
peerData.NextValidTime = nextTime
|
|
}
|
|
|
|
// RandomizeBackOff adds extra backoff period during which peer will not be dialed.
|
|
func (p *Status) RandomizeBackOff(pid peer.ID) {
|
|
p.store.Lock()
|
|
defer p.store.Unlock()
|
|
|
|
peerData := p.store.PeerDataGetOrCreate(pid)
|
|
|
|
// No need to add backoff period, if the previous one hasn't expired yet.
|
|
if !time.Now().After(peerData.NextValidTime) {
|
|
return
|
|
}
|
|
|
|
duration := time.Duration(math.Max(MinBackOffDuration, float64(p.rand.Intn(MaxBackOffDuration)))) * time.Millisecond
|
|
peerData.NextValidTime = time.Now().Add(duration)
|
|
}
|
|
|
|
// IsReadyToDial checks where the given peer is ready to be
|
|
// dialed again.
|
|
func (p *Status) IsReadyToDial(pid peer.ID) bool {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
|
|
if peerData, ok := p.store.PeerData(pid); ok {
|
|
timeIsZero := peerData.NextValidTime.IsZero()
|
|
isInvalidTime := peerData.NextValidTime.After(time.Now())
|
|
return timeIsZero || !isInvalidTime
|
|
}
|
|
// If no record exists, we don't restrict dials to the
|
|
// peer.
|
|
return true
|
|
}
|
|
|
|
// Connecting returns the peers that are connecting.
|
|
func (p *Status) Connecting() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
peers := make([]peer.ID, 0)
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.ConnState == PeerConnecting {
|
|
peers = append(peers, pid)
|
|
}
|
|
}
|
|
return peers
|
|
}
|
|
|
|
// Connected returns the peers that are connected.
|
|
func (p *Status) Connected() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
peers := make([]peer.ID, 0)
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.ConnState == PeerConnected {
|
|
peers = append(peers, pid)
|
|
}
|
|
}
|
|
return peers
|
|
}
|
|
|
|
// Inbound returns the current batch of inbound peers.
|
|
func (p *Status) Inbound() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
peers := make([]peer.ID, 0)
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.Direction == network.DirInbound {
|
|
peers = append(peers, pid)
|
|
}
|
|
}
|
|
return peers
|
|
}
|
|
|
|
// InboundConnected returns the current batch of inbound peers that are connected.
|
|
func (p *Status) InboundConnected() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
peers := make([]peer.ID, 0)
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.ConnState == PeerConnected && peerData.Direction == network.DirInbound {
|
|
peers = append(peers, pid)
|
|
}
|
|
}
|
|
return peers
|
|
}
|
|
|
|
// Outbound returns the current batch of outbound peers.
|
|
func (p *Status) Outbound() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
peers := make([]peer.ID, 0)
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.Direction == network.DirOutbound {
|
|
peers = append(peers, pid)
|
|
}
|
|
}
|
|
return peers
|
|
}
|
|
|
|
// OutboundConnected returns the current batch of outbound peers that are connected.
|
|
func (p *Status) OutboundConnected() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
peers := make([]peer.ID, 0)
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.ConnState == PeerConnected && peerData.Direction == network.DirOutbound {
|
|
peers = append(peers, pid)
|
|
}
|
|
}
|
|
return peers
|
|
}
|
|
|
|
// Active returns the peers that are connecting or connected.
|
|
func (p *Status) Active() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
peers := make([]peer.ID, 0)
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.ConnState == PeerConnecting || peerData.ConnState == PeerConnected {
|
|
peers = append(peers, pid)
|
|
}
|
|
}
|
|
return peers
|
|
}
|
|
|
|
// Disconnecting returns the peers that are disconnecting.
|
|
func (p *Status) Disconnecting() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
peers := make([]peer.ID, 0)
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.ConnState == PeerDisconnecting {
|
|
peers = append(peers, pid)
|
|
}
|
|
}
|
|
return peers
|
|
}
|
|
|
|
// Disconnected returns the peers that are disconnected.
|
|
func (p *Status) Disconnected() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
peers := make([]peer.ID, 0)
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.ConnState == PeerDisconnected {
|
|
peers = append(peers, pid)
|
|
}
|
|
}
|
|
return peers
|
|
}
|
|
|
|
// Inactive returns the peers that are disconnecting or disconnected.
|
|
func (p *Status) Inactive() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
peers := make([]peer.ID, 0)
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.ConnState == PeerDisconnecting || peerData.ConnState == PeerDisconnected {
|
|
peers = append(peers, pid)
|
|
}
|
|
}
|
|
return peers
|
|
}
|
|
|
|
// Bad returns the peers that are bad.
|
|
func (p *Status) Bad() []peer.ID {
|
|
return p.scorers.BadResponsesScorer().BadPeers()
|
|
}
|
|
|
|
// All returns all the peers regardless of state.
|
|
func (p *Status) All() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
pids := make([]peer.ID, 0, len(p.store.Peers()))
|
|
for pid := range p.store.Peers() {
|
|
pids = append(pids, pid)
|
|
}
|
|
return pids
|
|
}
|
|
|
|
// Prune clears out and removes outdated and disconnected peers.
|
|
func (p *Status) Prune() {
|
|
p.store.Lock()
|
|
defer p.store.Unlock()
|
|
|
|
// Default to old method if flag isnt enabled.
|
|
if !features.Get().EnablePeerScorer {
|
|
p.deprecatedPrune()
|
|
return
|
|
}
|
|
// Exit early if there is nothing to prune.
|
|
if len(p.store.Peers()) <= p.store.Config().MaxPeers {
|
|
return
|
|
}
|
|
notBadPeer := func(pid peer.ID) bool {
|
|
return !p.isBad(pid)
|
|
}
|
|
notTrustedPeer := func(pid peer.ID) bool {
|
|
return !p.isTrustedPeers(pid)
|
|
}
|
|
type peerResp struct {
|
|
pid peer.ID
|
|
score float64
|
|
}
|
|
peersToPrune := make([]*peerResp, 0)
|
|
// Select disconnected peers with a smaller bad response count.
|
|
for pid, peerData := range p.store.Peers() {
|
|
// Should not prune trusted peer or prune the peer dara and unset trusted peer.
|
|
if peerData.ConnState == PeerDisconnected && notBadPeer(pid) && notTrustedPeer(pid) {
|
|
peersToPrune = append(peersToPrune, &peerResp{
|
|
pid: pid,
|
|
score: p.Scorers().ScoreNoLock(pid),
|
|
})
|
|
}
|
|
}
|
|
|
|
// Sort peers in descending order, so the peers with the
|
|
// highest score are pruned first. This
|
|
// is to protect the node from malicious/lousy peers so
|
|
// that their memory is still kept.
|
|
sort.Slice(peersToPrune, func(i, j int) bool {
|
|
return peersToPrune[i].score > peersToPrune[j].score
|
|
})
|
|
|
|
limitDiff := len(p.store.Peers()) - p.store.Config().MaxPeers
|
|
if limitDiff > len(peersToPrune) {
|
|
limitDiff = len(peersToPrune)
|
|
}
|
|
|
|
peersToPrune = peersToPrune[:limitDiff]
|
|
|
|
// Delete peers from map.
|
|
for _, peerData := range peersToPrune {
|
|
p.store.DeletePeerData(peerData.pid)
|
|
}
|
|
p.tallyIPTracker()
|
|
}
|
|
|
|
// Deprecated: This is the old peer pruning method based on
|
|
// bad response counts.
|
|
func (p *Status) deprecatedPrune() {
|
|
// Exit early if there is nothing to prune.
|
|
if len(p.store.Peers()) <= p.store.Config().MaxPeers {
|
|
return
|
|
}
|
|
|
|
notBadPeer := func(peerData *peerdata.PeerData) bool {
|
|
return peerData.BadResponses < p.scorers.BadResponsesScorer().Params().Threshold
|
|
}
|
|
notTrustedPeer := func(pid peer.ID) bool {
|
|
return !p.isTrustedPeers(pid)
|
|
}
|
|
type peerResp struct {
|
|
pid peer.ID
|
|
badResp int
|
|
}
|
|
peersToPrune := make([]*peerResp, 0)
|
|
// Select disconnected peers with a smaller bad response count.
|
|
for pid, peerData := range p.store.Peers() {
|
|
// Should not prune trusted peer or prune the peer dara and unset trusted peer.
|
|
if peerData.ConnState == PeerDisconnected && notBadPeer(peerData) && notTrustedPeer(pid) {
|
|
peersToPrune = append(peersToPrune, &peerResp{
|
|
pid: pid,
|
|
badResp: peerData.BadResponses,
|
|
})
|
|
}
|
|
}
|
|
|
|
// Sort peers in ascending order, so the peers with the
|
|
// least amount of bad responses are pruned first. This
|
|
// is to protect the node from malicious/lousy peers so
|
|
// that their memory is still kept.
|
|
sort.Slice(peersToPrune, func(i, j int) bool {
|
|
return peersToPrune[i].badResp < peersToPrune[j].badResp
|
|
})
|
|
|
|
limitDiff := len(p.store.Peers()) - p.store.Config().MaxPeers
|
|
if limitDiff > len(peersToPrune) {
|
|
limitDiff = len(peersToPrune)
|
|
}
|
|
peersToPrune = peersToPrune[:limitDiff]
|
|
// Delete peers from map.
|
|
for _, peerData := range peersToPrune {
|
|
p.store.DeletePeerData(peerData.pid)
|
|
}
|
|
p.tallyIPTracker()
|
|
}
|
|
|
|
// BestFinalized returns the highest finalized epoch equal to or higher than ours that is agreed
|
|
// upon by the majority of peers. This method may not return the absolute highest finalized, but
|
|
// the finalized epoch in which most peers can serve blocks (plurality voting).
|
|
// Ideally, all peers would be reporting the same finalized epoch but some may be behind due to their
|
|
// own latency, or because of their finalized epoch at the time we queried them.
|
|
// Returns epoch number and list of peers that are at or beyond that epoch.
|
|
func (p *Status) BestFinalized(maxPeers int, ourFinalizedEpoch primitives.Epoch) (primitives.Epoch, []peer.ID) {
|
|
connected := p.Connected()
|
|
finalizedEpochVotes := make(map[primitives.Epoch]uint64)
|
|
pidEpoch := make(map[peer.ID]primitives.Epoch, len(connected))
|
|
pidHead := make(map[peer.ID]primitives.Slot, len(connected))
|
|
potentialPIDs := make([]peer.ID, 0, len(connected))
|
|
for _, pid := range connected {
|
|
peerChainState, err := p.ChainState(pid)
|
|
if err == nil && peerChainState != nil && peerChainState.FinalizedEpoch >= ourFinalizedEpoch {
|
|
finalizedEpochVotes[peerChainState.FinalizedEpoch]++
|
|
pidEpoch[pid] = peerChainState.FinalizedEpoch
|
|
potentialPIDs = append(potentialPIDs, pid)
|
|
pidHead[pid] = peerChainState.HeadSlot
|
|
}
|
|
}
|
|
|
|
// Select the target epoch, which is the epoch most peers agree upon.
|
|
var targetEpoch primitives.Epoch
|
|
var mostVotes uint64
|
|
for epoch, count := range finalizedEpochVotes {
|
|
if count > mostVotes || (count == mostVotes && epoch > targetEpoch) {
|
|
mostVotes = count
|
|
targetEpoch = epoch
|
|
}
|
|
}
|
|
|
|
// Sort PIDs by finalized epoch, in decreasing order.
|
|
sort.Slice(potentialPIDs, func(i, j int) bool {
|
|
if pidEpoch[potentialPIDs[i]] == pidEpoch[potentialPIDs[j]] {
|
|
return pidHead[potentialPIDs[i]] > pidHead[potentialPIDs[j]]
|
|
}
|
|
return pidEpoch[potentialPIDs[i]] > pidEpoch[potentialPIDs[j]]
|
|
})
|
|
|
|
// Trim potential peers to those on or after target epoch.
|
|
for i, pid := range potentialPIDs {
|
|
if pidEpoch[pid] < targetEpoch {
|
|
potentialPIDs = potentialPIDs[:i]
|
|
break
|
|
}
|
|
}
|
|
|
|
// Trim potential peers to at most maxPeers.
|
|
if len(potentialPIDs) > maxPeers {
|
|
potentialPIDs = potentialPIDs[:maxPeers]
|
|
}
|
|
|
|
return targetEpoch, potentialPIDs
|
|
}
|
|
|
|
// BestNonFinalized returns the highest known epoch, higher than ours,
|
|
// and is shared by at least minPeers.
|
|
func (p *Status) BestNonFinalized(minPeers int, ourHeadEpoch primitives.Epoch) (primitives.Epoch, []peer.ID) {
|
|
connected := p.Connected()
|
|
epochVotes := make(map[primitives.Epoch]uint64)
|
|
pidEpoch := make(map[peer.ID]primitives.Epoch, len(connected))
|
|
pidHead := make(map[peer.ID]primitives.Slot, len(connected))
|
|
potentialPIDs := make([]peer.ID, 0, len(connected))
|
|
|
|
ourHeadSlot := params.BeaconConfig().SlotsPerEpoch.Mul(uint64(ourHeadEpoch))
|
|
for _, pid := range connected {
|
|
peerChainState, err := p.ChainState(pid)
|
|
if err == nil && peerChainState != nil && peerChainState.HeadSlot > ourHeadSlot {
|
|
epoch := slots.ToEpoch(peerChainState.HeadSlot)
|
|
epochVotes[epoch]++
|
|
pidEpoch[pid] = epoch
|
|
pidHead[pid] = peerChainState.HeadSlot
|
|
potentialPIDs = append(potentialPIDs, pid)
|
|
}
|
|
}
|
|
|
|
// Select the target epoch, which has enough peers' votes (>= minPeers).
|
|
var targetEpoch primitives.Epoch
|
|
for epoch, votes := range epochVotes {
|
|
if votes >= uint64(minPeers) && targetEpoch < epoch {
|
|
targetEpoch = epoch
|
|
}
|
|
}
|
|
|
|
// Sort PIDs by head slot, in decreasing order.
|
|
sort.Slice(potentialPIDs, func(i, j int) bool {
|
|
return pidHead[potentialPIDs[i]] > pidHead[potentialPIDs[j]]
|
|
})
|
|
|
|
// Trim potential peers to those on or after target epoch.
|
|
for i, pid := range potentialPIDs {
|
|
if pidEpoch[pid] < targetEpoch {
|
|
potentialPIDs = potentialPIDs[:i]
|
|
break
|
|
}
|
|
}
|
|
|
|
return targetEpoch, potentialPIDs
|
|
}
|
|
|
|
// PeersToPrune selects the most suitable inbound peers
|
|
// to disconnect the host peer from. As of this moment
|
|
// the pruning relies on simple heuristics such as
|
|
// bad response count. In the future scoring will be used
|
|
// to determine the most suitable peers to take out.
|
|
func (p *Status) PeersToPrune() []peer.ID {
|
|
if !features.Get().EnablePeerScorer {
|
|
return p.deprecatedPeersToPrune()
|
|
}
|
|
connLimit := p.ConnectedPeerLimit()
|
|
inBoundLimit := uint64(p.InboundLimit())
|
|
activePeers := p.Active()
|
|
numInboundPeers := uint64(len(p.InboundConnected()))
|
|
// Exit early if we are still below our max
|
|
// limit.
|
|
if uint64(len(activePeers)) <= connLimit {
|
|
return []peer.ID{}
|
|
}
|
|
p.store.Lock()
|
|
defer p.store.Unlock()
|
|
|
|
type peerResp struct {
|
|
pid peer.ID
|
|
score float64
|
|
}
|
|
peersToPrune := make([]*peerResp, 0)
|
|
// Select connected and inbound peers to prune.
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.ConnState == PeerConnected &&
|
|
peerData.Direction == network.DirInbound && !p.store.IsTrustedPeer(pid) {
|
|
peersToPrune = append(peersToPrune, &peerResp{
|
|
pid: pid,
|
|
score: p.scorers.ScoreNoLock(pid),
|
|
})
|
|
}
|
|
}
|
|
|
|
// Sort in ascending order to favour pruning peers with a
|
|
// lower score.
|
|
sort.Slice(peersToPrune, func(i, j int) bool {
|
|
return peersToPrune[i].score < peersToPrune[j].score
|
|
})
|
|
|
|
// Determine amount of peers to prune using our
|
|
// max connection limit.
|
|
amountToPrune, err := pmath.Sub64(uint64(len(activePeers)), connLimit)
|
|
if err != nil {
|
|
// This should never happen.
|
|
log.WithError(err).Error("Failed to determine amount of peers to prune")
|
|
return []peer.ID{}
|
|
}
|
|
|
|
// Also check for inbound peers above our limit.
|
|
excessInbound := uint64(0)
|
|
if numInboundPeers > inBoundLimit {
|
|
excessInbound = numInboundPeers - inBoundLimit
|
|
}
|
|
// Prune the largest amount between excess peers and
|
|
// excess inbound peers.
|
|
if excessInbound > amountToPrune {
|
|
amountToPrune = excessInbound
|
|
}
|
|
if amountToPrune < uint64(len(peersToPrune)) {
|
|
peersToPrune = peersToPrune[:amountToPrune]
|
|
}
|
|
ids := make([]peer.ID, 0, len(peersToPrune))
|
|
for _, pr := range peersToPrune {
|
|
ids = append(ids, pr.pid)
|
|
}
|
|
return ids
|
|
}
|
|
|
|
// Deprecated: Is used to represent the older method
|
|
// of pruning which utilized bad response counts.
|
|
func (p *Status) deprecatedPeersToPrune() []peer.ID {
|
|
connLimit := p.ConnectedPeerLimit()
|
|
inBoundLimit := p.InboundLimit()
|
|
activePeers := p.Active()
|
|
numInboundPeers := len(p.InboundConnected())
|
|
// Exit early if we are still below our max
|
|
// limit.
|
|
if uint64(len(activePeers)) <= connLimit {
|
|
return []peer.ID{}
|
|
}
|
|
p.store.Lock()
|
|
defer p.store.Unlock()
|
|
|
|
type peerResp struct {
|
|
pid peer.ID
|
|
badResp int
|
|
}
|
|
peersToPrune := make([]*peerResp, 0)
|
|
// Select connected and inbound peers to prune.
|
|
for pid, peerData := range p.store.Peers() {
|
|
if peerData.ConnState == PeerConnected &&
|
|
peerData.Direction == network.DirInbound && !p.store.IsTrustedPeer(pid) {
|
|
peersToPrune = append(peersToPrune, &peerResp{
|
|
pid: pid,
|
|
badResp: peerData.BadResponses,
|
|
})
|
|
}
|
|
}
|
|
|
|
// Sort in descending order to favour pruning peers with a
|
|
// higher bad response count.
|
|
sort.Slice(peersToPrune, func(i, j int) bool {
|
|
return peersToPrune[i].badResp > peersToPrune[j].badResp
|
|
})
|
|
|
|
// Determine amount of peers to prune using our
|
|
// max connection limit.
|
|
amountToPrune, err := pmath.Sub64(uint64(len(activePeers)), connLimit)
|
|
if err != nil {
|
|
// This should never happen
|
|
log.WithError(err).Error("Failed to determine amount of peers to prune")
|
|
return []peer.ID{}
|
|
}
|
|
// Also check for inbound peers above our limit.
|
|
excessInbound := uint64(0)
|
|
if numInboundPeers > inBoundLimit {
|
|
excessInbound = uint64(numInboundPeers - inBoundLimit)
|
|
}
|
|
// Prune the largest amount between excess peers and
|
|
// excess inbound peers.
|
|
if excessInbound > amountToPrune {
|
|
amountToPrune = excessInbound
|
|
}
|
|
if amountToPrune < uint64(len(peersToPrune)) {
|
|
peersToPrune = peersToPrune[:amountToPrune]
|
|
}
|
|
ids := make([]peer.ID, 0, len(peersToPrune))
|
|
for _, pr := range peersToPrune {
|
|
ids = append(ids, pr.pid)
|
|
}
|
|
return ids
|
|
}
|
|
|
|
// HighestEpoch returns the highest epoch reported epoch amongst peers.
|
|
func (p *Status) HighestEpoch() primitives.Epoch {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
var highestSlot primitives.Slot
|
|
for _, peerData := range p.store.Peers() {
|
|
if peerData != nil && peerData.ChainState != nil && peerData.ChainState.HeadSlot > highestSlot {
|
|
highestSlot = peerData.ChainState.HeadSlot
|
|
}
|
|
}
|
|
return slots.ToEpoch(highestSlot)
|
|
}
|
|
|
|
// ConnectedPeerLimit returns the peer limit of
|
|
// concurrent peers connected to the beacon-node.
|
|
func (p *Status) ConnectedPeerLimit() uint64 {
|
|
maxLim := p.MaxPeerLimit()
|
|
if maxLim <= maxLimitBuffer {
|
|
return 0
|
|
}
|
|
return uint64(maxLim) - maxLimitBuffer
|
|
}
|
|
|
|
// SetTrustedPeers sets our trusted peer set into
|
|
// our peerstore.
|
|
func (p *Status) SetTrustedPeers(peers []peer.ID) {
|
|
p.store.Lock()
|
|
defer p.store.Unlock()
|
|
p.store.SetTrustedPeers(peers)
|
|
}
|
|
|
|
// GetTrustedPeers returns a list of all trusted peers' ids
|
|
func (p *Status) GetTrustedPeers() []peer.ID {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
return p.store.GetTrustedPeers()
|
|
}
|
|
|
|
// DeleteTrustedPeers removes peers from trusted peer set
|
|
func (p *Status) DeleteTrustedPeers(peers []peer.ID) {
|
|
p.store.Lock()
|
|
defer p.store.Unlock()
|
|
p.store.DeleteTrustedPeers(peers)
|
|
}
|
|
|
|
// IsTrustedPeers returns if given peer is a Trusted peer
|
|
func (p *Status) IsTrustedPeers(pid peer.ID) bool {
|
|
p.store.RLock()
|
|
defer p.store.RUnlock()
|
|
return p.isTrustedPeers(pid)
|
|
}
|
|
|
|
// isTrustedPeers is the lock-free version of IsTrustedPeers.
|
|
func (p *Status) isTrustedPeers(pid peer.ID) bool {
|
|
return p.store.IsTrustedPeer(pid)
|
|
}
|
|
|
|
// this method assumes the store lock is acquired before
|
|
// executing the method.
|
|
func (p *Status) isfromBadIP(pid peer.ID) bool {
|
|
peerData, ok := p.store.PeerData(pid)
|
|
if !ok {
|
|
return false
|
|
}
|
|
if peerData.Address == nil {
|
|
return false
|
|
}
|
|
ip, err := manet.ToIP(peerData.Address)
|
|
if err != nil {
|
|
return true
|
|
}
|
|
if val, ok := p.ipTracker[ip.String()]; ok {
|
|
if val > ColocationLimit {
|
|
return true
|
|
}
|
|
}
|
|
return false
|
|
}
|
|
|
|
func (p *Status) addIpToTracker(pid peer.ID) {
|
|
data, ok := p.store.PeerData(pid)
|
|
if !ok {
|
|
return
|
|
}
|
|
if data.Address == nil {
|
|
return
|
|
}
|
|
ip, err := manet.ToIP(data.Address)
|
|
if err != nil {
|
|
// Should never happen, it is
|
|
// assumed every IP coming in
|
|
// is a valid ip.
|
|
return
|
|
}
|
|
// Ignore loopback addresses.
|
|
if ip.IsLoopback() {
|
|
return
|
|
}
|
|
stringIP := ip.String()
|
|
p.ipTracker[stringIP] += 1
|
|
}
|
|
|
|
func (p *Status) tallyIPTracker() {
|
|
tracker := map[string]uint64{}
|
|
// Iterate through all peers.
|
|
for _, peerData := range p.store.Peers() {
|
|
if peerData.Address == nil {
|
|
continue
|
|
}
|
|
ip, err := manet.ToIP(peerData.Address)
|
|
if err != nil {
|
|
// Should never happen, it is
|
|
// assumed every IP coming in
|
|
// is a valid ip.
|
|
continue
|
|
}
|
|
stringIP := ip.String()
|
|
tracker[stringIP] += 1
|
|
}
|
|
p.ipTracker = tracker
|
|
}
|
|
|
|
func sameIP(firstAddr, secondAddr ma.Multiaddr) bool {
|
|
// Exit early if we do get nil multiaddresses
|
|
if firstAddr == nil || secondAddr == nil {
|
|
return false
|
|
}
|
|
firstIP, err := manet.ToIP(firstAddr)
|
|
if err != nil {
|
|
return false
|
|
}
|
|
secondIP, err := manet.ToIP(secondAddr)
|
|
if err != nil {
|
|
return false
|
|
}
|
|
return firstIP.Equal(secondIP)
|
|
}
|
|
|
|
func indicesFromBitfield(bitV bitfield.Bitvector64) []uint64 {
|
|
committeeIdxs := make([]uint64, 0, bitV.Count())
|
|
for i := uint64(0); i < 64; i++ {
|
|
if bitV.BitAt(i) {
|
|
committeeIdxs = append(committeeIdxs, i)
|
|
}
|
|
}
|
|
return committeeIdxs
|
|
}
|