Restart discovery when it fails

When discovery fails to be seeded with bootstrap/fallback nodes, it
never recovers.

This commit changes the behavior so that status-go retries fetching
bootnodes, and restarts discovery when that happens.
This commit is contained in:
Andrea Maria Piana 2022-12-09 16:16:21 +00:00
parent f98e65258a
commit 8593866862
28 changed files with 394 additions and 193 deletions

View File

@ -1 +1 @@
0.117.0
0.117.1

View File

@ -6,6 +6,7 @@ import (
"time"
"github.com/ethereum/go-ethereum/common"
"github.com/status-im/status-go/connection"
"github.com/status-im/status-go/eth-node/types"
"github.com/status-im/status-go/waku"
wakucommon "github.com/status-im/status-go/waku/common"
@ -256,6 +257,7 @@ func (w *gethWakuWrapper) MarkP2PMessageAsProcessed(hash common.Hash) {
func (w *gethWakuWrapper) RequestStoreMessages(peerID []byte, r types.MessagesRequest) (*types.StoreRequestCursor, error) {
return nil, errors.New("not implemented")
}
func (w *gethWakuWrapper) ConnectionChanged(_ connection.State) {}
type wakuFilterWrapper struct {
filter *wakucommon.Filter

View File

@ -10,6 +10,7 @@ import (
"github.com/waku-org/go-waku/waku/v2/protocol/store"
"github.com/ethereum/go-ethereum/common"
"github.com/status-im/status-go/connection"
"github.com/status-im/status-go/eth-node/types"
"github.com/status-im/status-go/wakuv2"
wakucommon "github.com/status-im/status-go/wakuv2/common"
@ -272,6 +273,10 @@ func (w *gethWakuV2Wrapper) SubscribeToConnStatusChanges() (*types.ConnStatusSub
return w.waku.SubscribeToConnStatusChanges(), nil
}
func (w *gethWakuV2Wrapper) ConnectionChanged(state connection.State) {
w.waku.ConnectionChanged(state)
}
type wakuV2FilterWrapper struct {
filter *wakucommon.Filter
id string

View File

@ -8,6 +8,7 @@ import (
"github.com/pborman/uuid"
"github.com/ethereum/go-ethereum/common"
"github.com/status-im/status-go/connection"
)
type ConnStatus struct {
@ -149,4 +150,7 @@ type Waku interface {
// MarkP2PMessageAsProcessed tells the waku layer that a P2P message has been processed
MarkP2PMessageAsProcessed(common.Hash)
// ConnectionChanged is called whenever the client knows its connection status has changed
ConnectionChanged(connection.State)
}

2
go.mod
View File

@ -80,7 +80,7 @@ require github.com/fogleman/gg v1.3.0
require (
github.com/gorilla/sessions v1.2.1
github.com/meirf/gopart v0.0.0-20180520194036-37e9492a85a8
github.com/waku-org/go-waku v0.2.3-test.0.20221209175307-685142e7b743
github.com/waku-org/go-waku v0.2.3-test.0.20221212154545-7443daea4cd4
)
require (

4
go.sum
View File

@ -2065,8 +2065,8 @@ github.com/vishvananda/netns v0.0.0-20200728191858-db3c7e526aae/go.mod h1:DD4vA1
github.com/vishvananda/netns v0.0.0-20210104183010-2eb08e3e575f/go.mod h1:DD4vA1DwXk04H54A1oHXtwZmA0grkVMdPxx/VGLCah0=
github.com/waku-org/go-discover v0.0.0-20221209174356-61c833f34d98 h1:xwY0kW5XZFimdqfZb9cZwT1S3VJP9j3AE6bdNd9boXM=
github.com/waku-org/go-discover v0.0.0-20221209174356-61c833f34d98/go.mod h1:eBHgM6T4EG0RZzxpxKy+rGz/6Dw2Nd8DWxS0lm9ESDw=
github.com/waku-org/go-waku v0.2.3-test.0.20221209175307-685142e7b743 h1:Q6bNLLCE7+OGrRlsmcrglKeURXmaagemIWTrvrJTgK4=
github.com/waku-org/go-waku v0.2.3-test.0.20221209175307-685142e7b743/go.mod h1:MzmxeUFKOSGqI+3ditwJVmiDXtWW7p4vZhmFeAcwKyI=
github.com/waku-org/go-waku v0.2.3-test.0.20221212154545-7443daea4cd4 h1:lzIDkSLbaAHzYlGgZB9BEWFEbVFZLmFAvzE2fiN61pg=
github.com/waku-org/go-waku v0.2.3-test.0.20221212154545-7443daea4cd4/go.mod h1:MzmxeUFKOSGqI+3ditwJVmiDXtWW7p4vZhmFeAcwKyI=
github.com/waku-org/go-zerokit-rln v0.1.7-wakuorg h1:2vVIBCtBih2w1K9ll8YnToTDZvbxcgbsClsPlJS/kkg=
github.com/waku-org/go-zerokit-rln v0.1.7-wakuorg/go.mod h1:GlyaVeEWNEBxVJrWC6jFTvb4LNb9d9qnjdS6EiWVUvk=
github.com/waku-org/noise v1.0.2 h1:7WmlhpJ0eliBzwzKz6SoTqQznaEU2IuebHF3oCekqqs=

View File

@ -643,7 +643,11 @@ func (s *MessengerContactRequestSuite) TestAcceptLatestContactRequestForContact(
resp, err = WaitOnMessengerResponse(
theirMessenger,
func(r *MessengerResponse) bool {
return len(r.Contacts) > 0 && len(r.Messages()) > 0 && len(r.ActivityCenterNotifications()) > 0
contactRequests, _, err := theirMessenger.PendingContactRequests("", 10)
if err != nil {
return false
}
return len(contactRequests) == 1
},
"no messages",
)

View File

@ -82,7 +82,9 @@ func (s *MessengerGroupChatSuite) createEmptyGroupChat(creator *Messenger, name
func (s *MessengerGroupChatSuite) verifyGroupChatCreated(member *Messenger, expectedChatActive bool) {
response, err := WaitOnMessengerResponse(
member,
func(r *MessengerResponse) bool { return len(r.Chats()) > 0 },
func(r *MessengerResponse) bool {
return len(r.Chats()) == 1 && r.Chats()[0].Active == expectedChatActive
},
"chat invitation not received",
)
s.Require().NoError(err)

View File

@ -800,6 +800,7 @@ func (m *Messenger) RemoveFilters(filters []*transport.Filter) error {
}
func (m *Messenger) ConnectionChanged(state connection.State) {
m.transport.ConnectionChanged(state)
if !m.connectionState.Offline && state.Offline {
m.sender.StopDatasync()
}

View File

@ -15,6 +15,7 @@ import (
"go.uber.org/zap"
"github.com/ethereum/go-ethereum/common"
"github.com/status-im/status-go/connection"
"github.com/status-im/status-go/eth-node/crypto"
"github.com/status-im/status-go/eth-node/types"
)
@ -79,8 +80,9 @@ type Transport struct {
// NewTransport returns a new Transport.
// TODO: leaving a chat should verify that for a given public key
// there are no other chats. It may happen that we leave a private chat
// but still have a public chat for a given public key.
//
// there are no other chats. It may happen that we leave a private chat
// but still have a public chat for a given public key.
func NewTransport(
waku types.Waku,
privateKey *ecdsa.PrivateKey,
@ -654,3 +656,7 @@ func (t *Transport) MarkP2PMessageAsProcessed(hash common.Hash) {
func (t *Transport) SubscribeToConnStatusChanges() (*types.ConnStatusSubscription, error) {
return t.waku.SubscribeToConnStatusChanges()
}
func (t *Transport) ConnectionChanged(state connection.State) {
t.waku.ConnectionChanged(state)
}

View File

@ -64,13 +64,28 @@ func WithDB(db *sql.DB) DBOption {
}
}
type ConnectionPoolOptions struct {
MaxOpenConnections int
MaxIdleConnections int
ConnectionMaxLifetime time.Duration
ConnectionMaxIdleTime time.Duration
}
// WithDriver is a DBOption that will open a *sql.DB connection
func WithDriver(driverName string, datasourceName string) DBOption {
func WithDriver(driverName string, datasourceName string, connectionPoolOptions ...ConnectionPoolOptions) DBOption {
return func(d *DBStore) error {
db, err := sql.Open(driverName, datasourceName)
if err != nil {
return err
}
if len(connectionPoolOptions) != 0 {
db.SetConnMaxIdleTime(connectionPoolOptions[0].ConnectionMaxIdleTime)
db.SetConnMaxLifetime(connectionPoolOptions[0].ConnectionMaxLifetime)
db.SetMaxIdleConns(connectionPoolOptions[0].MaxIdleConnections)
db.SetMaxOpenConns(connectionPoolOptions[0].MaxOpenConnections)
}
d.db = db
return nil
}
@ -119,31 +134,8 @@ func NewDBStore(log *zap.Logger, options ...DBOption) (*DBStore, error) {
}
}
// Disable concurrent access as not supported by the driver
result.db.SetMaxOpenConns(1)
var seq string
var name string
var file string // file will be empty if DB is :memory"
err := result.db.QueryRow("PRAGMA database_list").Scan(&seq, &name, &file)
if err != nil {
return nil, err
}
// readers do not block writers and faster i/o operations
// https://www.sqlite.org/draft/wal.html
// must be set after db is encrypted
var mode string
err = result.db.QueryRow("PRAGMA journal_mode=WAL").Scan(&mode)
if err != nil {
return nil, err
}
if mode != WALMode && file != "" {
return nil, fmt.Errorf("unable to set journal_mode to WAL. actual mode %s", mode)
}
if result.enableMigrations {
err = migrations.Migrate(result.db)
err := migrations.Migrate(result.db)
if err != nil {
return nil, err
}

View File

@ -8,9 +8,6 @@ import (
"sync"
"time"
"github.com/ethereum/go-ethereum/p2p/enode"
"github.com/ethereum/go-ethereum/p2p/enr"
"github.com/ethereum/go-ethereum/p2p/nat"
"github.com/libp2p/go-libp2p/core/discovery"
"github.com/libp2p/go-libp2p/core/host"
"github.com/libp2p/go-libp2p/core/peer"
@ -18,6 +15,10 @@ import (
"github.com/waku-org/go-waku/logging"
"github.com/waku-org/go-waku/waku/v2/utils"
"go.uber.org/zap"
"github.com/ethereum/go-ethereum/p2p/enode"
"github.com/ethereum/go-ethereum/p2p/enr"
"github.com/ethereum/go-ethereum/p2p/nat"
)
type DiscoveryV5 struct {
@ -26,23 +27,20 @@ type DiscoveryV5 struct {
discovery.Discovery
params *discV5Parameters
ctx context.Context
host host.Host
config discover.Config
udpAddr *net.UDPAddr
listener *discover.UDPv5
localnode *enode.LocalNode
NAT nat.Interface
quit chan struct{}
started bool
log *zap.Logger
wg *sync.WaitGroup
started bool
cancel context.CancelFunc
wg *sync.WaitGroup
peerCache peerCache
discoverCtx context.Context
discoverCancelFunc context.CancelFunc
peerCache peerCache
}
type peerCache struct {
@ -100,7 +98,7 @@ func DefaultOptions() []DiscoveryV5Option {
const MaxPeersToDiscover = 600
func NewDiscoveryV5(ctx context.Context, host host.Host, priv *ecdsa.PrivateKey, localnode *enode.LocalNode, log *zap.Logger, opts ...DiscoveryV5Option) (*DiscoveryV5, error) {
func NewDiscoveryV5(host host.Host, priv *ecdsa.PrivateKey, localnode *enode.LocalNode, log *zap.Logger, opts ...DiscoveryV5Option) (*DiscoveryV5, error) {
params := new(discV5Parameters)
optList := DefaultOptions()
optList = append(optList, opts...)
@ -116,7 +114,6 @@ func NewDiscoveryV5(ctx context.Context, host host.Host, priv *ecdsa.PrivateKey,
}
return &DiscoveryV5{
ctx: ctx,
host: host,
params: params,
NAT: NAT,
@ -149,7 +146,7 @@ func (d *DiscoveryV5) Node() *enode.Node {
return d.localnode.Node()
}
func (d *DiscoveryV5) listen() error {
func (d *DiscoveryV5) listen(ctx context.Context) error {
conn, err := net.ListenUDP("udp", d.udpAddr)
if err != nil {
return err
@ -160,7 +157,7 @@ func (d *DiscoveryV5) listen() error {
d.wg.Add(1)
go func() {
defer d.wg.Done()
nat.Map(d.NAT, d.quit, "udp", d.udpAddr.Port, d.udpAddr.Port, "go-waku discv5 discovery")
nat.Map(d.NAT, ctx.Done(), "udp", d.udpAddr.Port, d.udpAddr.Port, "go-waku discv5 discovery")
}()
}
@ -182,45 +179,35 @@ func (d *DiscoveryV5) listen() error {
return nil
}
func (d *DiscoveryV5) Start() error {
func (d *DiscoveryV5) Start(ctx context.Context) error {
d.Lock()
defer d.Unlock()
if d.started {
return nil
}
d.wg.Wait() // Waiting for any go routines to stop
ctx, cancel := context.WithCancel(ctx)
d.wg.Wait() // Waiting for other go routines to stop
d.quit = make(chan struct{}, 1)
d.cancel = cancel
d.started = true
err := d.listen()
err := d.listen(ctx)
if err != nil {
return err
}
// create cancellable
d.discoverCtx, d.discoverCancelFunc = context.WithCancel(d.ctx)
go d.runDiscoveryV5Loop()
go d.runDiscoveryV5Loop(ctx)
return nil
}
func (d *DiscoveryV5) SetBootnodes(nodes []*enode.Node) error {
return d.listener.SetFallbackNodes(nodes)
return d.listener.SetFallbackNodes(nodes)
}
func (d *DiscoveryV5) Stop() {
d.Lock()
defer d.Unlock()
if !d.started {
return
}
close(d.quit)
d.discoverCancelFunc()
d.cancel()
d.listener.Close()
d.listener = nil
@ -294,7 +281,7 @@ func (d *DiscoveryV5) Advertise(ctx context.Context, ns string, opts ...discover
return 20 * time.Minute, nil
}
func (d *DiscoveryV5) iterate(iterator enode.Iterator, limit int, doneCh chan struct{}) {
func (d *DiscoveryV5) iterate(ctx context.Context, iterator enode.Iterator, limit int) {
defer d.wg.Done()
for {
@ -302,7 +289,7 @@ func (d *DiscoveryV5) iterate(iterator enode.Iterator, limit int, doneCh chan st
break
}
if d.discoverCtx.Err() != nil {
if ctx.Err() != nil {
break
}
@ -333,8 +320,6 @@ func (d *DiscoveryV5) iterate(iterator enode.Iterator, limit int, doneCh chan st
}
d.peerCache.Unlock()
}
close(doneCh)
}
func (d *DiscoveryV5) removeExpiredPeers() int {
@ -353,21 +338,16 @@ func (d *DiscoveryV5) removeExpiredPeers() int {
return newCacheSize
}
func (d *DiscoveryV5) runDiscoveryV5Loop() {
func (d *DiscoveryV5) runDiscoveryV5Loop(ctx context.Context) {
iterator := d.listener.RandomNodes()
iterator = enode.Filter(iterator, evaluateNode)
defer iterator.Close()
doneCh := make(chan struct{})
d.wg.Add(1)
go d.iterate(iterator, MaxPeersToDiscover, doneCh)
go d.iterate(ctx, iterator, MaxPeersToDiscover)
select {
case <-d.discoverCtx.Done():
case <-doneCh:
}
<-ctx.Done()
d.log.Warn("Discv5 loop stopped")
}

View File

@ -96,7 +96,7 @@ func (w *WakuNode) connectednessListener() {
for {
select {
case <-w.quit:
case <-w.ctx.Done():
return
case <-w.protocolEventSub.Out():
case <-w.identificationEventSub.Out():

View File

@ -54,7 +54,8 @@ func (w *WakuNode) startKeepAlive(t time.Duration) {
}
lastTimeExecuted = w.timesource.Now()
case <-w.quit:
case <-w.ctx.Done():
w.log.Info("stopping ping protocol")
return
}
}

View File

@ -241,7 +241,7 @@ func (w *WakuNode) setupENR(addrs []ma.Multiaddr) error {
if w.discoveryV5 != nil && w.discoveryV5.IsStarted() {
w.log.Info("restarting discv5")
w.discoveryV5.Stop()
err = w.discoveryV5.Start()
err = w.discoveryV5.Start(w.ctx)
if err != nil {
w.log.Error("could not restart discv5", zap.Error(err))
return err

View File

@ -8,11 +8,12 @@ import (
"sync"
"time"
"github.com/libp2p/go-libp2p"
"go.uber.org/zap"
"github.com/ethereum/go-ethereum/common"
"github.com/ethereum/go-ethereum/crypto"
"github.com/ethereum/go-ethereum/p2p/enode"
"github.com/libp2p/go-libp2p"
"go.uber.org/zap"
pubsub "github.com/libp2p/go-libp2p-pubsub"
"github.com/libp2p/go-libp2p/core/event"
@ -94,9 +95,8 @@ type WakuNode struct {
keepAliveMutex sync.Mutex
keepAliveFails map[peer.ID]int
ctx context.Context
ctx context.Context // TODO: remove this
cancel context.CancelFunc
quit chan struct{}
wg *sync.WaitGroup
// Channel passed to WakuNode constructor
@ -170,7 +170,6 @@ func New(ctx context.Context, opts ...WakuNodeOption) (*WakuNode, error) {
w.ctx = ctx
w.opts = params
w.log = params.logger.Named("node2")
w.quit = make(chan struct{})
w.wg = &sync.WaitGroup{}
w.addrChan = make(chan ma.Multiaddr, 1024)
w.keepAliveFails = make(map[peer.ID]int)
@ -235,7 +234,7 @@ func (w *WakuNode) checkForAddressChanges() {
first <- struct{}{}
for {
select {
case <-w.quit:
case <-w.ctx.Done():
close(w.addrChan)
return
case <-first:
@ -268,7 +267,7 @@ func (w *WakuNode) checkForAddressChanges() {
// Start initializes all the protocols that were setup in the WakuNode
func (w *WakuNode) Start() error {
if w.opts.enableNTP {
err := w.timesource.Start()
err := w.timesource.Start(w.ctx)
if err != nil {
return err
}
@ -357,9 +356,7 @@ func (w *WakuNode) Start() error {
// Stop stops the WakuNode and closess all connections to the host
func (w *WakuNode) Stop() {
defer w.cancel()
close(w.quit)
w.cancel()
w.bcaster.Close()
@ -523,14 +520,14 @@ func (w *WakuNode) mountDiscV5() error {
}
var err error
w.discoveryV5, err = discv5.NewDiscoveryV5(w.ctx, w.Host(), w.opts.privKey, w.localNode, w.log, discV5Options...)
w.discoveryV5, err = discv5.NewDiscoveryV5(w.Host(), w.opts.privKey, w.localNode, w.log, discV5Options...)
return err
}
func (w *WakuNode) mountPeerExchange() error {
w.peerExchange = peer_exchange.NewWakuPeerExchange(w.ctx, w.host, w.discoveryV5, w.log)
return w.peerExchange.Start()
w.peerExchange = peer_exchange.NewWakuPeerExchange(w.host, w.discoveryV5, w.log)
return w.peerExchange.Start(w.ctx)
}
func (w *WakuNode) startStore() error {
@ -673,6 +670,12 @@ func (w *WakuNode) PeerStats() PeerStats {
return p
}
// Set the bootnodes on discv5
func (w *WakuNode) SetDiscV5Bootnodes(nodes []*enode.Node) error {
w.opts.discV5bootnodes = nodes
return w.discoveryV5.SetBootnodes(nodes)
}
// Peers return the list of peers, addresses, protocols supported and connection status
func (w *WakuNode) Peers() ([]*Peer, error) {
var peers []*Peer

View File

@ -4,7 +4,6 @@
package node
import (
"context"
"encoding/hex"
"errors"
@ -81,7 +80,7 @@ func (w *WakuNode) mountRlnRelay() error {
// mount the rln relay protocol in the on-chain/dynamic mode
var err error
w.rlnRelay, err = rln.RlnRelayDynamic(context.Background(), w.relay, w.opts.rlnETHClientAddress, w.opts.rlnETHPrivateKey, w.opts.rlnMembershipContractAddress, memKeyPair, w.opts.rlnRelayMemIndex, w.opts.rlnRelayPubsubTopic, w.opts.rlnRelayContentTopic, w.opts.rlnSpamHandler, w.opts.rlnRegistrationHandler, w.timesource, w.log)
w.rlnRelay, err = rln.RlnRelayDynamic(w.ctx, w.relay, w.opts.rlnETHClientAddress, w.opts.rlnETHPrivateKey, w.opts.rlnMembershipContractAddress, memKeyPair, w.opts.rlnRelayMemIndex, w.opts.rlnRelayPubsubTopic, w.opts.rlnRelayContentTopic, w.opts.rlnSpamHandler, w.opts.rlnRegistrationHandler, w.timesource, w.log)
if err != nil {
return err
}

View File

@ -317,8 +317,9 @@ func WithWakuFilter(fullNode bool, filterOpts ...filter.Option) WakuNodeOption {
}
// WithWakuStore enables the Waku V2 Store protocol and if the messages should
// be stored or not in a message provider
func WithWakuStore(shouldStoreMessages bool, resumeNodes []multiaddr.Multiaddr) WakuNodeOption {
// be stored or not in a message provider. If resumeNodes are specified, the
// store will attempt to resume message history using those nodes
func WithWakuStore(shouldStoreMessages bool, resumeNodes ...multiaddr.Multiaddr) WakuNodeOption {
return func(params *WakuNodeParameters) error {
params.enableStore = true
params.storeMsgs = shouldStoreMessages

View File

@ -46,24 +46,21 @@ type peerRecord struct {
type WakuPeerExchange struct {
h host.Host
ctx context.Context
disc *discv5.DiscoveryV5
log *zap.Logger
quit chan struct{}
wg sync.WaitGroup
log *zap.Logger
cancel context.CancelFunc
wg sync.WaitGroup
enrCache map[enode.ID]peerRecord // todo: next step: ring buffer; future: implement cache satisfying https://rfc.vac.dev/spec/34/
enrCacheMutex sync.RWMutex
rng *rand.Rand
started bool
}
// NewWakuPeerExchange returns a new instance of WakuPeerExchange struct
func NewWakuPeerExchange(ctx context.Context, h host.Host, disc *discv5.DiscoveryV5, log *zap.Logger) *WakuPeerExchange {
func NewWakuPeerExchange(h host.Host, disc *discv5.DiscoveryV5, log *zap.Logger) *WakuPeerExchange {
wakuPX := new(WakuPeerExchange)
wakuPX.ctx = ctx
wakuPX.h = h
wakuPX.disc = disc
wakuPX.log = log.Named("wakupx")
@ -73,19 +70,21 @@ func NewWakuPeerExchange(ctx context.Context, h host.Host, disc *discv5.Discover
}
// Start inits the peer exchange protocol
func (wakuPX *WakuPeerExchange) Start() error {
wakuPX.h.SetStreamHandlerMatch(PeerExchangeID_v20alpha1, protocol.PrefixTextMatch(string(PeerExchangeID_v20alpha1)), wakuPX.onRequest)
func (wakuPX *WakuPeerExchange) Start(ctx context.Context) error {
wakuPX.wg.Wait() // Waiting for any go routines to stop
ctx, cancel := context.WithCancel(ctx)
wakuPX.cancel = cancel
wakuPX.h.SetStreamHandlerMatch(PeerExchangeID_v20alpha1, protocol.PrefixTextMatch(string(PeerExchangeID_v20alpha1)), wakuPX.onRequest(ctx))
wakuPX.log.Info("Peer exchange protocol started")
wakuPX.started = true
wakuPX.quit = make(chan struct{}, 1)
wakuPX.wg.Add(1)
go wakuPX.runPeerExchangeDiscv5Loop()
go wakuPX.runPeerExchangeDiscv5Loop(ctx)
return nil
}
func (wakuPX *WakuPeerExchange) handleResponse(response *pb.PeerExchangeResponse) error {
func (wakuPX *WakuPeerExchange) handleResponse(ctx context.Context, response *pb.PeerExchangeResponse) error {
var peers []peer.AddrInfo
for _, p := range response.PeerInfos {
enrRecord := &enr.Record{}
@ -118,7 +117,7 @@ func (wakuPX *WakuPeerExchange) handleResponse(response *pb.PeerExchangeResponse
log.Info("connecting to newly discovered peers", zap.Int("count", len(peers)))
for _, p := range peers {
func(p peer.AddrInfo) {
ctx, cancel := context.WithTimeout(wakuPX.ctx, dialTimeout)
ctx, cancel := context.WithTimeout(ctx, dialTimeout)
defer cancel()
err := wakuPX.h.Connect(ctx, p)
if err != nil {
@ -131,35 +130,37 @@ func (wakuPX *WakuPeerExchange) handleResponse(response *pb.PeerExchangeResponse
return nil
}
func (wakuPX *WakuPeerExchange) onRequest(s network.Stream) {
defer s.Close()
logger := wakuPX.log.With(logging.HostID("peer", s.Conn().RemotePeer()))
requestRPC := &pb.PeerExchangeRPC{}
reader := protoio.NewDelimitedReader(s, math.MaxInt32)
err := reader.ReadMsg(requestRPC)
if err != nil {
logger.Error("reading request", zap.Error(err))
metrics.RecordPeerExchangeError(wakuPX.ctx, "decodeRpcFailure")
return
}
if requestRPC.Query != nil {
logger.Info("request received")
err := wakuPX.respond(requestRPC.Query.NumPeers, s.Conn().RemotePeer())
func (wakuPX *WakuPeerExchange) onRequest(ctx context.Context) func(s network.Stream) {
return func(s network.Stream) {
defer s.Close()
logger := wakuPX.log.With(logging.HostID("peer", s.Conn().RemotePeer()))
requestRPC := &pb.PeerExchangeRPC{}
reader := protoio.NewDelimitedReader(s, math.MaxInt32)
err := reader.ReadMsg(requestRPC)
if err != nil {
logger.Error("responding", zap.Error(err))
metrics.RecordPeerExchangeError(wakuPX.ctx, "pxFailure")
logger.Error("reading request", zap.Error(err))
metrics.RecordPeerExchangeError(ctx, "decodeRpcFailure")
return
}
}
if requestRPC.Response != nil {
logger.Info("response received")
err := wakuPX.handleResponse(requestRPC.Response)
if err != nil {
logger.Error("handling response", zap.Error(err))
metrics.RecordPeerExchangeError(wakuPX.ctx, "pxFailure")
return
if requestRPC.Query != nil {
logger.Info("request received")
err := wakuPX.respond(ctx, requestRPC.Query.NumPeers, s.Conn().RemotePeer())
if err != nil {
logger.Error("responding", zap.Error(err))
metrics.RecordPeerExchangeError(ctx, "pxFailure")
return
}
}
if requestRPC.Response != nil {
logger.Info("response received")
err := wakuPX.handleResponse(ctx, requestRPC.Response)
if err != nil {
logger.Error("handling response", zap.Error(err))
metrics.RecordPeerExchangeError(ctx, "pxFailure")
return
}
}
}
}
@ -176,7 +177,7 @@ func (wakuPX *WakuPeerExchange) Request(ctx context.Context, numPeers int, opts
}
if params.selectedPeer == "" {
metrics.RecordPeerExchangeError(wakuPX.ctx, "dialError")
metrics.RecordPeerExchangeError(ctx, "dialError")
return ErrNoPeersAvailable
}
@ -186,35 +187,27 @@ func (wakuPX *WakuPeerExchange) Request(ctx context.Context, numPeers int, opts
},
}
return wakuPX.sendPeerExchangeRPCToPeer(requestRPC, params.selectedPeer)
}
// IsStarted returns if the peer exchange protocol has been mounted or not
func (wakuPX *WakuPeerExchange) IsStarted() bool {
return wakuPX.started
return wakuPX.sendPeerExchangeRPCToPeer(ctx, requestRPC, params.selectedPeer)
}
// Stop unmounts the peer exchange protocol
func (wakuPX *WakuPeerExchange) Stop() {
if wakuPX.started {
wakuPX.h.RemoveStreamHandler(PeerExchangeID_v20alpha1)
wakuPX.started = false
close(wakuPX.quit)
wakuPX.wg.Wait()
}
wakuPX.cancel()
wakuPX.h.RemoveStreamHandler(PeerExchangeID_v20alpha1)
wakuPX.wg.Wait()
}
func (wakuPX *WakuPeerExchange) sendPeerExchangeRPCToPeer(rpc *pb.PeerExchangeRPC, peerID peer.ID) error {
func (wakuPX *WakuPeerExchange) sendPeerExchangeRPCToPeer(ctx context.Context, rpc *pb.PeerExchangeRPC, peerID peer.ID) error {
logger := wakuPX.log.With(logging.HostID("peer", peerID))
// We connect first so dns4 addresses are resolved (NewStream does not do it)
err := wakuPX.h.Connect(wakuPX.ctx, wakuPX.h.Peerstore().PeerInfo(peerID))
err := wakuPX.h.Connect(ctx, wakuPX.h.Peerstore().PeerInfo(peerID))
if err != nil {
logger.Error("connecting peer", zap.Error(err))
return err
}
connOpt, err := wakuPX.h.NewStream(wakuPX.ctx, peerID, PeerExchangeID_v20alpha1)
connOpt, err := wakuPX.h.NewStream(ctx, peerID, PeerExchangeID_v20alpha1)
if err != nil {
logger.Error("creating stream to peer", zap.Error(err))
return err
@ -231,7 +224,7 @@ func (wakuPX *WakuPeerExchange) sendPeerExchangeRPCToPeer(rpc *pb.PeerExchangeRP
return nil
}
func (wakuPX *WakuPeerExchange) respond(numPeers uint64, peerID peer.ID) error {
func (wakuPX *WakuPeerExchange) respond(ctx context.Context, numPeers uint64, peerID peer.ID) error {
records, err := wakuPX.getENRsFromCache(numPeers)
if err != nil {
return err
@ -241,7 +234,7 @@ func (wakuPX *WakuPeerExchange) respond(numPeers uint64, peerID peer.ID) error {
responseRPC.Response = new(pb.PeerExchangeResponse)
responseRPC.Response.PeerInfos = records
return wakuPX.sendPeerExchangeRPCToPeer(responseRPC, peerID)
return wakuPX.sendPeerExchangeRPCToPeer(ctx, responseRPC, peerID)
}
func (wakuPX *WakuPeerExchange) getENRsFromCache(numPeers uint64) ([]*pb.PeerInfo, error) {
@ -304,12 +297,8 @@ func (wakuPX *WakuPeerExchange) cleanCache() {
wakuPX.enrCache = r
}
func (wakuPX *WakuPeerExchange) findPeers() {
if !wakuPX.disc.IsStarted() {
return
}
ctx, cancel := context.WithTimeout(wakuPX.ctx, 2*time.Second)
func (wakuPX *WakuPeerExchange) findPeers(ctx context.Context) {
ctx, cancel := context.WithTimeout(ctx, 2*time.Second)
defer cancel()
peerRecords, err := wakuPX.disc.FindNodes(ctx, "")
if err != nil {
@ -332,7 +321,7 @@ func (wakuPX *WakuPeerExchange) findPeers() {
wakuPX.cleanCache()
}
func (wakuPX *WakuPeerExchange) runPeerExchangeDiscv5Loop() {
func (wakuPX *WakuPeerExchange) runPeerExchangeDiscv5Loop(ctx context.Context) {
defer wakuPX.wg.Done()
// Runs a discv5 loop adding new peers to the px peer cache
@ -349,15 +338,15 @@ func (wakuPX *WakuPeerExchange) runPeerExchangeDiscv5Loop() {
// This loop "competes" with the loop in wakunode2
// For the purpose of collecting px peers, 30 sec intervals should be enough
wakuPX.findPeers()
wakuPX.findPeers(ctx)
for {
select {
case <-wakuPX.quit:
case <-ctx.Done():
return
case <-ticker.C:
wakuPX.findPeers()
wakuPX.findPeers(ctx)
}
}

View File

@ -247,7 +247,7 @@ func (w *WakuRelay) SubscribeToTopic(ctx context.Context, topic string) (*Subscr
w.bcaster.Register(&topic, subscription.C)
}
go w.subscribeToTopic(topic, subscription, sub)
go w.subscribeToTopic(ctx, topic, subscription, sub)
return subscription, nil
}
@ -307,8 +307,8 @@ func (w *WakuRelay) nextMessage(ctx context.Context, sub *pubsub.Subscription) <
return msgChannel
}
func (w *WakuRelay) subscribeToTopic(t string, subscription *Subscription, sub *pubsub.Subscription) {
ctx, err := tag.New(context.Background(), tag.Insert(metrics.KeyType, "relay"))
func (w *WakuRelay) subscribeToTopic(ctx context.Context, t string, subscription *Subscription, sub *pubsub.Subscription) {
ctx, err := tag.New(ctx, tag.Insert(metrics.KeyType, "relay"))
if err != nil {
w.log.Error("creating tag map", zap.Error(err))
return

View File

@ -33,7 +33,7 @@ func register(ctx context.Context, idComm r.IDCommitment, ethAccountPrivateKey *
}
defer backend.Close()
chainID, err := backend.ChainID(context.Background())
chainID, err := backend.ChainID(ctx)
if err != nil {
return nil, err
}

View File

@ -256,10 +256,13 @@ func (store *WakuStore) Query(ctx context.Context, query Query, opts ...HistoryR
result := &Result{
Messages: response.Messages,
query: q,
cursor: response.PagingInfo.Cursor,
peerId: params.selectedPeer,
}
if response.PagingInfo != nil {
result.cursor = response.PagingInfo.Cursor
}
return result, nil
}

View File

@ -2,6 +2,7 @@ package timesource
import (
"bytes"
"context"
"errors"
"sort"
"sync"
@ -133,8 +134,8 @@ type NTPTimeSource struct {
timeQuery ntpQuery // for ease of testing
log *zap.Logger
quit chan struct{}
wg sync.WaitGroup
cancel context.CancelFunc
wg sync.WaitGroup
mu sync.RWMutex
latestOffset time.Duration
@ -162,9 +163,11 @@ func (s *NTPTimeSource) updateOffset() error {
// runPeriodically runs periodically the given function based on NTPTimeSource
// synchronization limits (fastNTPSyncPeriod / slowNTPSyncPeriod)
func (s *NTPTimeSource) runPeriodically(fn func() error) error {
func (s *NTPTimeSource) runPeriodically(ctx context.Context, fn func() error) error {
var period time.Duration
s.quit = make(chan struct{})
s.log.Info("starting service")
// we try to do it synchronously so that user can have reliable messages right away
s.wg.Add(1)
go func() {
@ -177,7 +180,8 @@ func (s *NTPTimeSource) runPeriodically(fn func() error) error {
period = s.fastNTPSyncPeriod
}
case <-s.quit:
case <-ctx.Done():
s.log.Info("stopping service")
s.wg.Done()
return
}
@ -188,16 +192,16 @@ func (s *NTPTimeSource) runPeriodically(fn func() error) error {
}
// Start runs a goroutine that updates local offset every updatePeriod.
func (s *NTPTimeSource) Start() error {
return s.runPeriodically(s.updateOffset)
func (s *NTPTimeSource) Start(ctx context.Context) error {
s.wg.Wait() // Waiting for other go routines to stop
ctx, cancel := context.WithCancel(ctx)
s.cancel = cancel
return s.runPeriodically(ctx, s.updateOffset)
}
// Stop goroutine that updates time source.
func (s *NTPTimeSource) Stop() error {
if s.quit == nil {
return nil
}
close(s.quit)
s.cancel()
s.wg.Wait()
return nil
}

View File

@ -1,9 +1,12 @@
package timesource
import "time"
import (
"context"
"time"
)
type Timesource interface {
Now() time.Time
Start() error
Start(ctx context.Context) error
Stop() error
}

View File

@ -1,6 +1,9 @@
package timesource
import "time"
import (
"context"
"time"
)
type WallClockTimeSource struct {
}
@ -13,7 +16,7 @@ func (t *WallClockTimeSource) Now() time.Time {
return time.Now()
}
func (t *WallClockTimeSource) Start() error {
func (t *WallClockTimeSource) Start(ctx context.Context) error {
// Do nothing
return nil
}

2
vendor/modules.txt vendored
View File

@ -990,7 +990,7 @@ github.com/vacp2p/mvds/transport
github.com/waku-org/go-discover/discover
github.com/waku-org/go-discover/discover/v4wire
github.com/waku-org/go-discover/discover/v5wire
# github.com/waku-org/go-waku v0.2.3-test.0.20221209175307-685142e7b743
# github.com/waku-org/go-waku v0.2.3-test.0.20221212154545-7443daea4cd4
## explicit; go 1.18
github.com/waku-org/go-waku/logging
github.com/waku-org/go-waku/waku/persistence

View File

@ -25,6 +25,7 @@ import (
"database/sql"
"errors"
"fmt"
"math"
"math/rand"
"net"
"runtime"
@ -62,6 +63,7 @@ import (
"github.com/waku-org/go-waku/waku/v2/protocol/relay"
"github.com/waku-org/go-waku/waku/v2/utils"
"github.com/status-im/status-go/connection"
"github.com/status-im/status-go/eth-node/types"
"github.com/status-im/status-go/signal"
"github.com/status-im/status-go/timesource"
@ -77,6 +79,8 @@ import (
const messageQueueLimit = 1024
const requestTimeout = 30 * time.Second
const autoRelayMinInterval = 2 * time.Second
const bootnodesQueryBackoffMs = 200
const bootnodesMaxRetries = 7
type settings struct {
LightClient bool // Indicates if the node is a light client
@ -132,12 +136,29 @@ type Waku struct {
// NTP Synced timesource
timesource *timesource.NTPTimeSource
// seededBootnodesForDiscV5 indicates whether we manage to retrieve discovery
// bootnodes successfully
seededBootnodesForDiscV5 bool
// offline indicates whether we have detected connectivity
offline bool
// connectionChanged is channel that notifies when connectivity has changed
connectionChanged chan struct{}
// discV5BootstrapNodes is the ENR to be used to fetch bootstrap nodes for discovery
discV5BootstrapNodes []string
}
// New creates a WakuV2 client ready to communicate through the LibP2P network.
func New(nodeKey string, fleet string, cfg *Config, logger *zap.Logger, appDB *sql.DB, timesource *timesource.NTPTimeSource) (*Waku, error) {
var err error
if logger == nil {
logger = zap.NewNop()
logger, err = zap.NewDevelopment()
if err != nil {
return nil, err
}
}
cfg = setDefaults(cfg)
@ -154,6 +175,7 @@ func New(nodeKey string, fleet string, cfg *Config, logger *zap.Logger, appDB *s
sendQueue: make(chan *pb.WakuMessage, 1000),
connStatusSubscriptions: make(map[string]*types.ConnStatusSubscription),
quit: make(chan struct{}),
connectionChanged: make(chan struct{}),
wg: sync.WaitGroup{},
dnsAddressCache: make(map[string][]dnsdisc.DiscoveredNode),
dnsAddressCacheLock: &sync.RWMutex{},
@ -161,6 +183,7 @@ func New(nodeKey string, fleet string, cfg *Config, logger *zap.Logger, appDB *s
storeMsgIDsMu: sync.RWMutex{},
timeSource: time.Now,
logger: logger,
discV5BootstrapNodes: cfg.DiscV5BootstrapNodes,
}
// Disabling light client mode if using status.prod or undefined
@ -181,7 +204,6 @@ func New(nodeKey string, fleet string, cfg *Config, logger *zap.Logger, appDB *s
waku.filterMsgChannel = make(chan *protocol.Envelope, 1024)
var privateKey *ecdsa.PrivateKey
var err error
if nodeKey != "" {
privateKey, err = crypto.HexToECDSA(nodeKey)
} else {
@ -226,8 +248,10 @@ func New(nodeKey string, fleet string, cfg *Config, logger *zap.Logger, appDB *s
if cfg.EnableDiscV5 {
bootnodes, err := waku.getDiscV5BootstrapNodes(ctx, cfg.DiscV5BootstrapNodes)
if err != nil {
logger.Error("failed to get bootstrap nodes", zap.Error(err))
return nil, err
}
opts = append(opts, node.WithDiscoveryV5(cfg.UDPPort, bootnodes, cfg.AutoUpdate, pubsub.WithDiscoveryOpts(discovery.Limit(cfg.DiscoveryLimit))))
// Peer exchange requires DiscV5 to run (might change in future versions of the protocol)
@ -276,7 +300,7 @@ func New(nodeKey string, fleet string, cfg *Config, logger *zap.Logger, appDB *s
}
if cfg.EnableDiscV5 {
err := waku.node.DiscV5().Start()
err := waku.node.DiscV5().Start(ctx)
if err != nil {
return nil, err
}
@ -358,12 +382,16 @@ func (w *Waku) getDiscV5BootstrapNodes(ctx context.Context, addresses []string)
}
}
wg.Wait()
w.seededBootnodesForDiscV5 = len(result) > 0
return result, nil
}
type fnApplyToEachPeer func(d dnsdisc.DiscoveredNode, wg *sync.WaitGroup)
func (w *Waku) dnsDiscover(ctx context.Context, enrtreeAddress string, apply fnApplyToEachPeer) {
w.logger.Info("retrieving nodes", zap.String("enr", enrtreeAddress))
ctx, cancel := context.WithTimeout(ctx, requestTimeout)
defer cancel()
@ -372,15 +400,17 @@ func (w *Waku) dnsDiscover(ctx context.Context, enrtreeAddress string, apply fnA
discNodes, ok := w.dnsAddressCache[enrtreeAddress]
if !ok {
// NOTE: Temporary fix for DNS resolution on android/ios, as gomobile does not support it
// NOTE: Temporary fix for DNS resolution on android/ios, as gomobile does not support it
discoveredNodes, err := dnsdisc.RetrieveNodes(ctx, enrtreeAddress, dnsdisc.WithNameserver("1.1.1.1"))
if err != nil {
w.logger.Warn("dns discovery error ", zap.Error(err))
return
}
w.dnsAddressCache[enrtreeAddress] = append(w.dnsAddressCache[enrtreeAddress], discoveredNodes...)
discNodes = w.dnsAddressCache[enrtreeAddress]
if len(discoveredNodes) != 0 {
w.dnsAddressCache[enrtreeAddress] = append(w.dnsAddressCache[enrtreeAddress], discoveredNodes...)
discNodes = w.dnsAddressCache[enrtreeAddress]
}
}
wg := &sync.WaitGroup{}
@ -1093,6 +1123,7 @@ func (w *Waku) Start() error {
}
go w.broadcast()
go w.seedBootnodesForDiscV5()
return nil
}
@ -1104,6 +1135,7 @@ func (w *Waku) Stop() error {
w.node.Stop()
close(w.quit)
close(w.filterMsgChannel)
close(w.connectionChanged)
w.wg.Wait()
return nil
}
@ -1256,7 +1288,7 @@ func (w *Waku) StartDiscV5() error {
return errors.New("discv5 is not setup")
}
return w.node.DiscV5().Start()
return w.node.DiscV5().Start(context.Background())
}
func (w *Waku) StopDiscV5() error {
@ -1268,6 +1300,88 @@ func (w *Waku) StopDiscV5() error {
return nil
}
func (w *Waku) ConnectionChanged(state connection.State) {
if !state.Offline && w.offline {
select {
case w.connectionChanged <- struct{}{}:
}
}
w.offline = !state.Offline
}
// seedBootnodesForDiscV5 tries to fetch bootnodes
// from an ENR periodically.
// It backs off exponentially until maxRetries, at which point it restarts from 0
// It also restarts if there's a connection change signalled from the client
func (w *Waku) seedBootnodesForDiscV5() {
ticker := time.NewTicker(200 * time.Millisecond)
defer ticker.Stop()
var lastTry = time.Now().UnixNano() / int64(time.Millisecond)
var retries = 0
for {
select {
case <-ticker.C:
if w.seededBootnodesForDiscV5 {
w.logger.Info("stopped querying bootnodes")
return
}
now := time.Now().UnixNano() / int64(time.Millisecond)
backoff := bootnodesQueryBackoffMs * int64(math.Exp2(float64(retries)))
if lastTry+backoff < now {
err := w.restartDiscV5()
if err != nil {
w.logger.Warn("failed to restart discv5", zap.Error(err))
}
lastTry = now
retries++
// We reset the retries after a while and restart
if retries > bootnodesMaxRetries {
retries = 0
}
}
// If we go online, trigger immediately
case <-w.connectionChanged:
now := time.Now().UnixNano() / int64(time.Millisecond)
backoff := bootnodesQueryBackoffMs * int64(math.Exp2(float64(retries)))
// check we haven't run too eagerly, in case connection
// is flapping
if lastTry+backoff < now {
err := w.restartDiscV5()
if err != nil {
w.logger.Warn("failed to restart discv5", zap.Error(err))
}
}
retries = 0
lastTry = now
case <-w.quit:
return
}
}
}
// Restart discv5, re-retrieving bootstrap nodes
func (w *Waku) restartDiscV5() error {
ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
defer cancel()
bootnodes, err := w.getDiscV5BootstrapNodes(ctx, w.discV5BootstrapNodes)
if err != nil {
return err
}
if len(bootnodes) == 0 {
return errors.New("failed to fetch bootnodes")
}
w.logger.Info("restarting discv5 with nodes", zap.Any("nodes", bootnodes))
return w.node.SetDiscV5Bootnodes(bootnodes)
}
func (w *Waku) AddStorePeer(address string) (string, error) {
addr, err := multiaddr.NewMultiaddr(address)
if err != nil {

85
wakuv2/waku_test.go Normal file
View File

@ -0,0 +1,85 @@
package wakuv2
import (
"errors"
"testing"
"time"
"github.com/cenkalti/backoff/v3"
"github.com/stretchr/testify/require"
"github.com/status-im/status-go/protocol/tt"
)
var testENRBootstrap = "enrtree://AOGECG2SPND25EEFMAJ5WF3KSGJNSGV356DSTL2YVLLZWIV6SAYBM@prod.nodes.status.im"
func TestDiscoveryV5(t *testing.T) {
config := &Config{}
config.EnableDiscV5 = true
config.DiscV5BootstrapNodes = []string{testENRBootstrap}
config.DiscoveryLimit = 20
config.UDPPort = 9001
w, err := New("", "", config, nil, nil, nil)
require.NoError(t, err)
require.NoError(t, w.Start())
err = tt.RetryWithBackOff(func() error {
if len(w.Peers()) == 0 {
return errors.New("no peers discovered")
}
return nil
})
require.NoError(t, err)
require.NotEqual(t, 0, len(w.Peers()))
require.NoError(t, w.Stop())
}
func TestRestartDiscoveryV5(t *testing.T) {
config := &Config{}
config.EnableDiscV5 = true
// Use wrong discv5 bootstrap address, to simulate being offline
config.DiscV5BootstrapNodes = []string{"enrtree://AOGECG2SPND25EEFMAJ5WF3KSGJNSGV356DSTL2YVLLZWIV6SAYBM@1.1.1.2"}
config.DiscoveryLimit = 20
config.UDPPort = 9002
w, err := New("", "", config, nil, nil, nil)
require.NoError(t, err)
require.NoError(t, w.Start())
require.False(t, w.seededBootnodesForDiscV5)
options := func(b *backoff.ExponentialBackOff) {
b.MaxElapsedTime = 2 * time.Second
}
// Sanity check, not great, but it's probably helpful
err = tt.RetryWithBackOff(func() error {
if len(w.Peers()) == 0 {
return errors.New("no peers discovered")
}
return nil
}, options)
require.Error(t, err)
w.discV5BootstrapNodes = []string{testENRBootstrap}
options = func(b *backoff.ExponentialBackOff) {
b.MaxElapsedTime = 30 * time.Second
}
err = tt.RetryWithBackOff(func() error {
if len(w.Peers()) == 0 {
return errors.New("no peers discovered")
}
return nil
}, options)
require.NoError(t, err)
require.True(t, w.seededBootnodesForDiscV5)
require.NotEqual(t, 0, len(w.Peers()))
require.NoError(t, w.Stop())
}