go-ethereum/p2p/simulations/http_test.go

871 lines
21 KiB
Go
Raw Normal View History

// Copyright 2017 The go-ethereum Authors
// This file is part of the go-ethereum library.
//
// The go-ethereum library is free software: you can redistribute it and/or modify
// it under the terms of the GNU Lesser General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// The go-ethereum library is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Lesser General Public License for more details.
//
// You should have received a copy of the GNU Lesser General Public License
// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
package simulations
import (
"context"
"flag"
"fmt"
"math/rand"
"net/http/httptest"
"os"
"reflect"
"sync"
"sync/atomic"
"testing"
"time"
"github.com/ethereum/go-ethereum/event"
"github.com/ethereum/go-ethereum/log"
"github.com/ethereum/go-ethereum/node"
"github.com/ethereum/go-ethereum/p2p"
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 17:59:00 -05:00
"github.com/ethereum/go-ethereum/p2p/enode"
"github.com/ethereum/go-ethereum/p2p/simulations/adapters"
"github.com/ethereum/go-ethereum/rpc"
"github.com/mattn/go-colorable"
)
func TestMain(m *testing.M) {
loglevel := flag.Int("loglevel", 2, "verbosity of logs")
flag.Parse()
log.PrintOrigins(true)
log.Root().SetHandler(log.LvlFilterHandler(log.Lvl(*loglevel), log.StreamHandler(colorable.NewColorableStderr(), log.TerminalFormat(true))))
os.Exit(m.Run())
}
// testService implements the node.Service interface and provides protocols
// and APIs which are useful for testing nodes in a simulation network
type testService struct {
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 17:59:00 -05:00
id enode.ID
// peerCount is incremented once a peer handshake has been performed
peerCount int64
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 17:59:00 -05:00
peers map[enode.ID]*testPeer
peersMtx sync.Mutex
// state stores []byte which is used to test creating and loading
// snapshots
state atomic.Value
}
func newTestService(ctx *adapters.ServiceContext) (node.Service, error) {
svc := &testService{
id: ctx.Config.ID,
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 17:59:00 -05:00
peers: make(map[enode.ID]*testPeer),
}
svc.state.Store(ctx.Snapshot)
return svc, nil
}
type testPeer struct {
testReady chan struct{}
dumReady chan struct{}
}
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 17:59:00 -05:00
func (t *testService) peer(id enode.ID) *testPeer {
t.peersMtx.Lock()
defer t.peersMtx.Unlock()
if peer, ok := t.peers[id]; ok {
return peer
}
peer := &testPeer{
testReady: make(chan struct{}),
dumReady: make(chan struct{}),
}
t.peers[id] = peer
return peer
}
func (t *testService) Protocols() []p2p.Protocol {
return []p2p.Protocol{
{
Name: "test",
Version: 1,
Length: 3,
Run: t.RunTest,
},
{
Name: "dum",
Version: 1,
Length: 1,
Run: t.RunDum,
},
{
Name: "prb",
Version: 1,
Length: 1,
Run: t.RunPrb,
},
}
}
func (t *testService) APIs() []rpc.API {
return []rpc.API{{
Namespace: "test",
Version: "1.0",
Service: &TestAPI{
state: &t.state,
peerCount: &t.peerCount,
},
}}
}
func (t *testService) Start(server *p2p.Server) error {
return nil
}
func (t *testService) Stop() error {
return nil
}
// handshake performs a peer handshake by sending and expecting an empty
// message with the given code
func (t *testService) handshake(rw p2p.MsgReadWriter, code uint64) error {
errc := make(chan error, 2)
go func() { errc <- p2p.Send(rw, code, struct{}{}) }()
go func() { errc <- p2p.ExpectMsg(rw, code, struct{}{}) }()
for i := 0; i < 2; i++ {
if err := <-errc; err != nil {
return err
}
}
return nil
}
func (t *testService) RunTest(p *p2p.Peer, rw p2p.MsgReadWriter) error {
peer := t.peer(p.ID())
// perform three handshakes with three different message codes,
// used to test message sending and filtering
if err := t.handshake(rw, 2); err != nil {
return err
}
if err := t.handshake(rw, 1); err != nil {
return err
}
if err := t.handshake(rw, 0); err != nil {
return err
}
// close the testReady channel so that other protocols can run
close(peer.testReady)
// track the peer
atomic.AddInt64(&t.peerCount, 1)
defer atomic.AddInt64(&t.peerCount, -1)
// block until the peer is dropped
for {
_, err := rw.ReadMsg()
if err != nil {
return err
}
}
}
func (t *testService) RunDum(p *p2p.Peer, rw p2p.MsgReadWriter) error {
peer := t.peer(p.ID())
// wait for the test protocol to perform its handshake
<-peer.testReady
// perform a handshake
if err := t.handshake(rw, 0); err != nil {
return err
}
// close the dumReady channel so that other protocols can run
close(peer.dumReady)
// block until the peer is dropped
for {
_, err := rw.ReadMsg()
if err != nil {
return err
}
}
}
func (t *testService) RunPrb(p *p2p.Peer, rw p2p.MsgReadWriter) error {
peer := t.peer(p.ID())
// wait for the dum protocol to perform its handshake
<-peer.dumReady
// perform a handshake
if err := t.handshake(rw, 0); err != nil {
return err
}
// block until the peer is dropped
for {
_, err := rw.ReadMsg()
if err != nil {
return err
}
}
}
func (t *testService) Snapshot() ([]byte, error) {
return t.state.Load().([]byte), nil
}
// TestAPI provides a test API to:
// * get the peer count
// * get and set an arbitrary state byte slice
// * get and increment a counter
// * subscribe to counter increment events
type TestAPI struct {
state *atomic.Value
peerCount *int64
counter int64
feed event.Feed
}
func (t *TestAPI) PeerCount() int64 {
return atomic.LoadInt64(t.peerCount)
}
func (t *TestAPI) Get() int64 {
return atomic.LoadInt64(&t.counter)
}
func (t *TestAPI) Add(delta int64) {
atomic.AddInt64(&t.counter, delta)
t.feed.Send(delta)
}
func (t *TestAPI) GetState() []byte {
return t.state.Load().([]byte)
}
func (t *TestAPI) SetState(state []byte) {
t.state.Store(state)
}
func (t *TestAPI) Events(ctx context.Context) (*rpc.Subscription, error) {
notifier, supported := rpc.NotifierFromContext(ctx)
if !supported {
return nil, rpc.ErrNotificationsUnsupported
}
rpcSub := notifier.CreateSubscription()
go func() {
events := make(chan int64)
sub := t.feed.Subscribe(events)
defer sub.Unsubscribe()
for {
select {
case event := <-events:
notifier.Notify(rpcSub.ID, event)
case <-sub.Err():
return
case <-rpcSub.Err():
return
case <-notifier.Closed():
return
}
}
}()
return rpcSub, nil
}
var testServices = adapters.Services{
"test": newTestService,
}
func testHTTPServer(t *testing.T) (*Network, *httptest.Server) {
t.Helper()
adapter := adapters.NewSimAdapter(testServices)
network := NewNetwork(adapter, &NetworkConfig{
DefaultService: "test",
})
return network, httptest.NewServer(NewServer(network))
}
// TestHTTPNetwork tests interacting with a simulation network using the HTTP
// API
func TestHTTPNetwork(t *testing.T) {
// start the server
network, s := testHTTPServer(t)
defer s.Close()
// subscribe to events so we can check them later
client := NewClient(s.URL)
events := make(chan *Event, 100)
var opts SubscribeOpts
sub, err := client.SubscribeNetwork(events, opts)
if err != nil {
t.Fatalf("error subscribing to network events: %s", err)
}
defer sub.Unsubscribe()
// check we can retrieve details about the network
gotNetwork, err := client.GetNetwork()
if err != nil {
t.Fatalf("error getting network: %s", err)
}
if gotNetwork.ID != network.ID {
t.Fatalf("expected network to have ID %q, got %q", network.ID, gotNetwork.ID)
}
// start a simulation network
nodeIDs := startTestNetwork(t, client)
// check we got all the events
x := &expectEvents{t, events, sub}
x.expect(
x.nodeEvent(nodeIDs[0], false),
x.nodeEvent(nodeIDs[1], false),
x.nodeEvent(nodeIDs[0], true),
x.nodeEvent(nodeIDs[1], true),
x.connEvent(nodeIDs[0], nodeIDs[1], false),
x.connEvent(nodeIDs[0], nodeIDs[1], true),
)
// reconnect the stream and check we get the current nodes and conns
events = make(chan *Event, 100)
opts.Current = true
sub, err = client.SubscribeNetwork(events, opts)
if err != nil {
t.Fatalf("error subscribing to network events: %s", err)
}
defer sub.Unsubscribe()
x = &expectEvents{t, events, sub}
x.expect(
x.nodeEvent(nodeIDs[0], true),
x.nodeEvent(nodeIDs[1], true),
x.connEvent(nodeIDs[0], nodeIDs[1], true),
)
}
func startTestNetwork(t *testing.T, client *Client) []string {
// create two nodes
nodeCount := 2
nodeIDs := make([]string, nodeCount)
for i := 0; i < nodeCount; i++ {
config := adapters.RandomNodeConfig()
node, err := client.CreateNode(config)
if err != nil {
t.Fatalf("error creating node: %s", err)
}
nodeIDs[i] = node.ID
}
// check both nodes exist
nodes, err := client.GetNodes()
if err != nil {
t.Fatalf("error getting nodes: %s", err)
}
if len(nodes) != nodeCount {
t.Fatalf("expected %d nodes, got %d", nodeCount, len(nodes))
}
for i, nodeID := range nodeIDs {
if nodes[i].ID != nodeID {
t.Fatalf("expected node %d to have ID %q, got %q", i, nodeID, nodes[i].ID)
}
node, err := client.GetNode(nodeID)
if err != nil {
t.Fatalf("error getting node %d: %s", i, err)
}
if node.ID != nodeID {
t.Fatalf("expected node %d to have ID %q, got %q", i, nodeID, node.ID)
}
}
// start both nodes
for _, nodeID := range nodeIDs {
if err := client.StartNode(nodeID); err != nil {
t.Fatalf("error starting node %q: %s", nodeID, err)
}
}
// connect the nodes
for i := 0; i < nodeCount-1; i++ {
peerId := i + 1
if i == nodeCount-1 {
peerId = 0
}
if err := client.ConnectNode(nodeIDs[i], nodeIDs[peerId]); err != nil {
t.Fatalf("error connecting nodes: %s", err)
}
}
return nodeIDs
}
type expectEvents struct {
*testing.T
events chan *Event
sub event.Subscription
}
func (t *expectEvents) nodeEvent(id string, up bool) *Event {
config := &adapters.NodeConfig{ID: enode.HexID(id)}
return &Event{Type: EventTypeNode, Node: newNode(nil, config, up)}
}
func (t *expectEvents) connEvent(one, other string, up bool) *Event {
return &Event{
Type: EventTypeConn,
Conn: &Conn{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 17:59:00 -05:00
One: enode.HexID(one),
Other: enode.HexID(other),
Up: up,
},
}
}
func (t *expectEvents) expectMsgs(expected map[MsgFilter]int) {
actual := make(map[MsgFilter]int)
timeout := time.After(10 * time.Second)
loop:
for {
select {
case event := <-t.events:
t.Logf("received %s event: %v", event.Type, event)
if event.Type != EventTypeMsg || event.Msg.Received {
continue loop
}
if event.Msg == nil {
t.Fatal("expected event.Msg to be set")
}
filter := MsgFilter{
Proto: event.Msg.Protocol,
Code: int64(event.Msg.Code),
}
actual[filter]++
if actual[filter] > expected[filter] {
t.Fatalf("received too many msgs for filter: %v", filter)
}
if reflect.DeepEqual(actual, expected) {
return
}
case err := <-t.sub.Err():
t.Fatalf("network stream closed unexpectedly: %s", err)
case <-timeout:
t.Fatal("timed out waiting for expected events")
}
}
}
func (t *expectEvents) expect(events ...*Event) {
p2p, swarm: fix node up races by granular locking (#18976) * swarm/network: DRY out repeated giga comment I not necessarily agree with the way we wait for event propagation. But I truly disagree with having duplicated giga comments. * p2p/simulations: encapsulate Node.Up field so we avoid data races The Node.Up field was accessed concurrently without "proper" locking. There was a lock on Network and that was used sometimes to access the field. Other times the locking was missed and we had a data race. For example: https://github.com/ethereum/go-ethereum/pull/18464 The case above was solved, but there were still intermittent/hard to reproduce races. So let's solve the issue permanently. resolves: ethersphere/go-ethereum#1146 * p2p/simulations: fix unmarshal of simulations.Node Making Node.Up field private in 13292ee897e345045fbfab3bda23a77589a271c1 broke TestHTTPNetwork and TestHTTPSnapshot. Because the default UnmarshalJSON does not handle unexported fields. Important: The fix is partial and not proper to my taste. But I cut scope as I think the fix may require a change to the current serialization format. New ticket: https://github.com/ethersphere/go-ethereum/issues/1177 * p2p/simulations: Add a sanity test case for Node.Config UnmarshalJSON * p2p/simulations: revert back to defer Unlock() pattern for Network It's a good patten to call `defer Unlock()` right after `Lock()` so (new) error cases won't miss to unlock. Let's get back to that pattern. The patten was abandoned in 85a79b3ad3c5863f8612d25c246bcfad339f36b7, while fixing a data race. That data race does not exist anymore, since the Node.Up field got hidden behind its own lock. * p2p/simulations: consistent naming for test providers Node.UnmarshalJSON * p2p/simulations: remove JSON annotation from private fields of Node As unexported fields are not serialized. * p2p/simulations: fix deadlock in Network.GetRandomDownNode() Problem: GetRandomDownNode() locks -> getDownNodeIDs() -> GetNodes() tries to lock -> deadlock On Network type, unexported functions must assume that `net.lock` is already acquired and should not call exported functions which might try to lock again. * p2p/simulations: ensure method conformity for Network Connect* methods were moved to p2p/simulations.Network from swarm/network/simulation. However these new methods did not follow the pattern of Network methods, i.e., all exported method locks the whole Network either for read or write. * p2p/simulations: fix deadlock during network shutdown `TestDiscoveryPersistenceSimulationSimAdapter` often got into deadlock. The execution was stuck on two locks, i.e, `Kademlia.lock` and `p2p/simulations.Network.lock`. Usually the test got stuck once in each 20 executions with high confidence. `Kademlia` was stuck in `Kademlia.EachAddr()` and `Network` in `Network.Stop()`. Solution: in `Network.Stop()` `net.lock` must be released before calling `node.Stop()` as stopping a node (somehow - I did not find the exact code path) causes `Network.InitConn()` to be called from `Kademlia.SuggestPeer()` and that blocks on `net.lock`. Related ticket: https://github.com/ethersphere/go-ethereum/issues/1223 * swarm/state: simplify if statement in DBStore.Put() * p2p/simulations: remove faulty godoc from private function The comment started with the wrong method name. The method is simple and self explanatory. Also, it's private. => Let's just remove the comment.
2019-02-18 00:38:14 -06:00
t.Helper()
timeout := time.After(10 * time.Second)
i := 0
for {
select {
case event := <-t.events:
t.Logf("received %s event: %v", event.Type, event)
expected := events[i]
if event.Type != expected.Type {
t.Fatalf("expected event %d to have type %q, got %q", i, expected.Type, event.Type)
}
switch expected.Type {
case EventTypeNode:
if event.Node == nil {
t.Fatal("expected event.Node to be set")
}
if event.Node.ID() != expected.Node.ID() {
t.Fatalf("expected node event %d to have id %q, got %q", i, expected.Node.ID().TerminalString(), event.Node.ID().TerminalString())
}
p2p, swarm: fix node up races by granular locking (#18976) * swarm/network: DRY out repeated giga comment I not necessarily agree with the way we wait for event propagation. But I truly disagree with having duplicated giga comments. * p2p/simulations: encapsulate Node.Up field so we avoid data races The Node.Up field was accessed concurrently without "proper" locking. There was a lock on Network and that was used sometimes to access the field. Other times the locking was missed and we had a data race. For example: https://github.com/ethereum/go-ethereum/pull/18464 The case above was solved, but there were still intermittent/hard to reproduce races. So let's solve the issue permanently. resolves: ethersphere/go-ethereum#1146 * p2p/simulations: fix unmarshal of simulations.Node Making Node.Up field private in 13292ee897e345045fbfab3bda23a77589a271c1 broke TestHTTPNetwork and TestHTTPSnapshot. Because the default UnmarshalJSON does not handle unexported fields. Important: The fix is partial and not proper to my taste. But I cut scope as I think the fix may require a change to the current serialization format. New ticket: https://github.com/ethersphere/go-ethereum/issues/1177 * p2p/simulations: Add a sanity test case for Node.Config UnmarshalJSON * p2p/simulations: revert back to defer Unlock() pattern for Network It's a good patten to call `defer Unlock()` right after `Lock()` so (new) error cases won't miss to unlock. Let's get back to that pattern. The patten was abandoned in 85a79b3ad3c5863f8612d25c246bcfad339f36b7, while fixing a data race. That data race does not exist anymore, since the Node.Up field got hidden behind its own lock. * p2p/simulations: consistent naming for test providers Node.UnmarshalJSON * p2p/simulations: remove JSON annotation from private fields of Node As unexported fields are not serialized. * p2p/simulations: fix deadlock in Network.GetRandomDownNode() Problem: GetRandomDownNode() locks -> getDownNodeIDs() -> GetNodes() tries to lock -> deadlock On Network type, unexported functions must assume that `net.lock` is already acquired and should not call exported functions which might try to lock again. * p2p/simulations: ensure method conformity for Network Connect* methods were moved to p2p/simulations.Network from swarm/network/simulation. However these new methods did not follow the pattern of Network methods, i.e., all exported method locks the whole Network either for read or write. * p2p/simulations: fix deadlock during network shutdown `TestDiscoveryPersistenceSimulationSimAdapter` often got into deadlock. The execution was stuck on two locks, i.e, `Kademlia.lock` and `p2p/simulations.Network.lock`. Usually the test got stuck once in each 20 executions with high confidence. `Kademlia` was stuck in `Kademlia.EachAddr()` and `Network` in `Network.Stop()`. Solution: in `Network.Stop()` `net.lock` must be released before calling `node.Stop()` as stopping a node (somehow - I did not find the exact code path) causes `Network.InitConn()` to be called from `Kademlia.SuggestPeer()` and that blocks on `net.lock`. Related ticket: https://github.com/ethersphere/go-ethereum/issues/1223 * swarm/state: simplify if statement in DBStore.Put() * p2p/simulations: remove faulty godoc from private function The comment started with the wrong method name. The method is simple and self explanatory. Also, it's private. => Let's just remove the comment.
2019-02-18 00:38:14 -06:00
if event.Node.Up() != expected.Node.Up() {
t.Fatalf("expected node event %d to have up=%t, got up=%t", i, expected.Node.Up(), event.Node.Up())
}
case EventTypeConn:
if event.Conn == nil {
t.Fatal("expected event.Conn to be set")
}
if event.Conn.One != expected.Conn.One {
t.Fatalf("expected conn event %d to have one=%q, got one=%q", i, expected.Conn.One.TerminalString(), event.Conn.One.TerminalString())
}
if event.Conn.Other != expected.Conn.Other {
t.Fatalf("expected conn event %d to have other=%q, got other=%q", i, expected.Conn.Other.TerminalString(), event.Conn.Other.TerminalString())
}
if event.Conn.Up != expected.Conn.Up {
t.Fatalf("expected conn event %d to have up=%t, got up=%t", i, expected.Conn.Up, event.Conn.Up)
}
}
i++
if i == len(events) {
return
}
case err := <-t.sub.Err():
t.Fatalf("network stream closed unexpectedly: %s", err)
case <-timeout:
t.Fatal("timed out waiting for expected events")
}
}
}
// TestHTTPNodeRPC tests calling RPC methods on nodes via the HTTP API
func TestHTTPNodeRPC(t *testing.T) {
// start the server
_, s := testHTTPServer(t)
defer s.Close()
// start a node in the network
client := NewClient(s.URL)
config := adapters.RandomNodeConfig()
node, err := client.CreateNode(config)
if err != nil {
t.Fatalf("error creating node: %s", err)
}
if err := client.StartNode(node.ID); err != nil {
t.Fatalf("error starting node: %s", err)
}
// create two RPC clients
ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
defer cancel()
rpcClient1, err := client.RPCClient(ctx, node.ID)
if err != nil {
t.Fatalf("error getting node RPC client: %s", err)
}
rpcClient2, err := client.RPCClient(ctx, node.ID)
if err != nil {
t.Fatalf("error getting node RPC client: %s", err)
}
// subscribe to events using client 1
events := make(chan int64, 1)
sub, err := rpcClient1.Subscribe(ctx, "test", events, "events")
if err != nil {
t.Fatalf("error subscribing to events: %s", err)
}
defer sub.Unsubscribe()
// call some RPC methods using client 2
if err := rpcClient2.CallContext(ctx, nil, "test_add", 10); err != nil {
t.Fatalf("error calling RPC method: %s", err)
}
var result int64
if err := rpcClient2.CallContext(ctx, &result, "test_get"); err != nil {
t.Fatalf("error calling RPC method: %s", err)
}
if result != 10 {
t.Fatalf("expected result to be 10, got %d", result)
}
// check we got an event from client 1
select {
case event := <-events:
if event != 10 {
t.Fatalf("expected event to be 10, got %d", event)
}
case <-ctx.Done():
t.Fatal(ctx.Err())
}
}
// TestHTTPSnapshot tests creating and loading network snapshots
func TestHTTPSnapshot(t *testing.T) {
// start the server
network, s := testHTTPServer(t)
defer s.Close()
var eventsDone = make(chan struct{})
count := 1
eventsDoneChan := make(chan *Event)
eventSub := network.Events().Subscribe(eventsDoneChan)
go func() {
defer eventSub.Unsubscribe()
for event := range eventsDoneChan {
if event.Type == EventTypeConn && !event.Control {
count--
if count == 0 {
eventsDone <- struct{}{}
return
}
}
}
}()
// create a two-node network
client := NewClient(s.URL)
nodeCount := 2
nodes := make([]*p2p.NodeInfo, nodeCount)
for i := 0; i < nodeCount; i++ {
config := adapters.RandomNodeConfig()
node, err := client.CreateNode(config)
if err != nil {
t.Fatalf("error creating node: %s", err)
}
if err := client.StartNode(node.ID); err != nil {
t.Fatalf("error starting node: %s", err)
}
nodes[i] = node
}
if err := client.ConnectNode(nodes[0].ID, nodes[1].ID); err != nil {
t.Fatalf("error connecting nodes: %s", err)
}
// store some state in the test services
states := make([]string, nodeCount)
for i, node := range nodes {
rpc, err := client.RPCClient(context.Background(), node.ID)
if err != nil {
t.Fatalf("error getting RPC client: %s", err)
}
defer rpc.Close()
state := fmt.Sprintf("%x", rand.Int())
if err := rpc.Call(nil, "test_setState", []byte(state)); err != nil {
t.Fatalf("error setting service state: %s", err)
}
states[i] = state
}
<-eventsDone
// create a snapshot
snap, err := client.CreateSnapshot()
if err != nil {
t.Fatalf("error creating snapshot: %s", err)
}
for i, state := range states {
gotState := snap.Nodes[i].Snapshots["test"]
if string(gotState) != state {
t.Fatalf("expected snapshot state %q, got %q", state, gotState)
}
}
// create another network
network2, s := testHTTPServer(t)
defer s.Close()
client = NewClient(s.URL)
count = 1
eventSub = network2.Events().Subscribe(eventsDoneChan)
go func() {
defer eventSub.Unsubscribe()
for event := range eventsDoneChan {
if event.Type == EventTypeConn && !event.Control {
count--
if count == 0 {
eventsDone <- struct{}{}
return
}
}
}
}()
// subscribe to events so we can check them later
events := make(chan *Event, 100)
var opts SubscribeOpts
sub, err := client.SubscribeNetwork(events, opts)
if err != nil {
t.Fatalf("error subscribing to network events: %s", err)
}
defer sub.Unsubscribe()
// load the snapshot
if err := client.LoadSnapshot(snap); err != nil {
t.Fatalf("error loading snapshot: %s", err)
}
<-eventsDone
// check the nodes and connection exists
net, err := client.GetNetwork()
if err != nil {
t.Fatalf("error getting network: %s", err)
}
if len(net.Nodes) != nodeCount {
t.Fatalf("expected network to have %d nodes, got %d", nodeCount, len(net.Nodes))
}
for i, node := range nodes {
id := net.Nodes[i].ID().String()
if id != node.ID {
t.Fatalf("expected node %d to have ID %s, got %s", i, node.ID, id)
}
}
if len(net.Conns) != 1 {
t.Fatalf("expected network to have 1 connection, got %d", len(net.Conns))
}
conn := net.Conns[0]
if conn.One.String() != nodes[0].ID {
t.Fatalf("expected connection to have one=%q, got one=%q", nodes[0].ID, conn.One)
}
if conn.Other.String() != nodes[1].ID {
t.Fatalf("expected connection to have other=%q, got other=%q", nodes[1].ID, conn.Other)
}
if !conn.Up {
t.Fatal("should be up")
}
// check the node states were restored
for i, node := range nodes {
rpc, err := client.RPCClient(context.Background(), node.ID)
if err != nil {
t.Fatalf("error getting RPC client: %s", err)
}
defer rpc.Close()
var state []byte
if err := rpc.Call(&state, "test_getState"); err != nil {
t.Fatalf("error getting service state: %s", err)
}
if string(state) != states[i] {
t.Fatalf("expected snapshot state %q, got %q", states[i], state)
}
}
// check we got all the events
x := &expectEvents{t, events, sub}
x.expect(
x.nodeEvent(nodes[0].ID, false),
x.nodeEvent(nodes[0].ID, true),
x.nodeEvent(nodes[1].ID, false),
x.nodeEvent(nodes[1].ID, true),
x.connEvent(nodes[0].ID, nodes[1].ID, false),
x.connEvent(nodes[0].ID, nodes[1].ID, true),
)
}
// TestMsgFilterPassMultiple tests streaming message events using a filter
// with multiple protocols
func TestMsgFilterPassMultiple(t *testing.T) {
// start the server
_, s := testHTTPServer(t)
defer s.Close()
// subscribe to events with a message filter
client := NewClient(s.URL)
events := make(chan *Event, 10)
opts := SubscribeOpts{
Filter: "prb:0-test:0",
}
sub, err := client.SubscribeNetwork(events, opts)
if err != nil {
t.Fatalf("error subscribing to network events: %s", err)
}
defer sub.Unsubscribe()
// start a simulation network
startTestNetwork(t, client)
// check we got the expected events
x := &expectEvents{t, events, sub}
x.expectMsgs(map[MsgFilter]int{
{"test", 0}: 2,
{"prb", 0}: 2,
})
}
// TestMsgFilterPassWildcard tests streaming message events using a filter
// with a code wildcard
func TestMsgFilterPassWildcard(t *testing.T) {
// start the server
_, s := testHTTPServer(t)
defer s.Close()
// subscribe to events with a message filter
client := NewClient(s.URL)
events := make(chan *Event, 10)
opts := SubscribeOpts{
Filter: "prb:0,2-test:*",
}
sub, err := client.SubscribeNetwork(events, opts)
if err != nil {
t.Fatalf("error subscribing to network events: %s", err)
}
defer sub.Unsubscribe()
// start a simulation network
startTestNetwork(t, client)
// check we got the expected events
x := &expectEvents{t, events, sub}
x.expectMsgs(map[MsgFilter]int{
{"test", 2}: 2,
{"test", 1}: 2,
{"test", 0}: 2,
{"prb", 0}: 2,
})
}
// TestMsgFilterPassSingle tests streaming message events using a filter
// with a single protocol and code
func TestMsgFilterPassSingle(t *testing.T) {
// start the server
_, s := testHTTPServer(t)
defer s.Close()
// subscribe to events with a message filter
client := NewClient(s.URL)
events := make(chan *Event, 10)
opts := SubscribeOpts{
Filter: "dum:0",
}
sub, err := client.SubscribeNetwork(events, opts)
if err != nil {
t.Fatalf("error subscribing to network events: %s", err)
}
defer sub.Unsubscribe()
// start a simulation network
startTestNetwork(t, client)
// check we got the expected events
x := &expectEvents{t, events, sub}
x.expectMsgs(map[MsgFilter]int{
{"dum", 0}: 2,
})
}
// TestMsgFilterPassSingle tests streaming message events using an invalid
// filter
func TestMsgFilterFailBadParams(t *testing.T) {
// start the server
_, s := testHTTPServer(t)
defer s.Close()
client := NewClient(s.URL)
events := make(chan *Event, 10)
opts := SubscribeOpts{
Filter: "foo:",
}
_, err := client.SubscribeNetwork(events, opts)
if err == nil {
t.Fatalf("expected event subscription to fail but succeeded!")
}
opts.Filter = "bzz:aa"
_, err = client.SubscribeNetwork(events, opts)
if err == nil {
t.Fatalf("expected event subscription to fail but succeeded!")
}
opts.Filter = "invalid"
_, err = client.SubscribeNetwork(events, opts)
if err == nil {
t.Fatalf("expected event subscription to fail but succeeded!")
}
}