op-geth/swarm/network/stream/syncer_test.go

360 lines
10 KiB
Go
Raw Normal View History

2018-06-20 12:06:27 +00:00
// Copyright 2018 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 stream
import (
"context"
"errors"
2018-06-20 12:06:27 +00:00
"fmt"
"io/ioutil"
"math"
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
"os"
2018-06-20 12:06:27 +00:00
"sync"
"testing"
"time"
"github.com/ethereum/go-ethereum/common"
"github.com/ethereum/go-ethereum/node"
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 22:59:00 +00:00
"github.com/ethereum/go-ethereum/p2p/enode"
"github.com/ethereum/go-ethereum/p2p/simulations/adapters"
2018-06-20 12:06:27 +00:00
"github.com/ethereum/go-ethereum/swarm/log"
"github.com/ethereum/go-ethereum/swarm/network"
"github.com/ethereum/go-ethereum/swarm/network/simulation"
"github.com/ethereum/go-ethereum/swarm/state"
2018-06-20 12:06:27 +00:00
"github.com/ethereum/go-ethereum/swarm/storage"
"github.com/ethereum/go-ethereum/swarm/storage/mock"
"github.com/ethereum/go-ethereum/swarm/testutil"
2018-06-20 12:06:27 +00:00
)
const dataChunkCount = 200
func TestSyncerSimulation(t *testing.T) {
testSyncBetweenNodes(t, 2, dataChunkCount, true, 1)
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
// This test uses much more memory when running with
// race detector. Allow it to finish successfully by
// reducing its scope, and still check for data races
// with the smallest number of nodes.
Ci race detector handle failing tests (#19143) * swarm/storage: increase mget timeout in common_test.go TestDbStoreCorrect_1k sometimes timed out with -race on Travis. --- FAIL: TestDbStoreCorrect_1k (24.63s) common_test.go:194: testStore failed: timed out after 10s * swarm: remove unused vars from TestSnapshotSyncWithServer nodeCount and chunkCount is returned from setupSim and those values we use. * swarm: move race/norace helpers from stream to testutil As we will need to use the flag in other packages, too. * swarm: refactor TestSwarmNetwork case Extract long running test cases for better visibility. * swarm/network: skip TestSyncingViaGlobalSync with -race As panics on Travis. panic: runtime error: invalid memory address or nil pointer dereference [signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0x7e351b] * swarm: run TestSwarmNetwork with fewer nodes with -race As otherwise we always get test failure with `network_test.go:374: context deadline exceeded` even with raised `Timeout`. * swarm/network: run TestDeliveryFromNodes with fewer nodes with -race Test on Travis times out with 8 or more nodes if -race flag is present. * swarm/network: smaller node count for discovery tests with -race TestDiscoveryPersistenceSimulationSimAdapters failed on Travis with `-race` flag present. The failure was due to extensive memory usage, coming from the CGO runtime. Using a smaller node count resolves the issue. === RUN TestDiscoveryPersistenceSimulationSimAdapter ==7227==ERROR: ThreadSanitizer failed to allocate 0x80000 (524288) bytes of clock allocator (error code: 12) FATAL: ThreadSanitizer CHECK failed: ./gotsan.cc:6976 "((0 && "unable to mmap")) != (0)" (0x0, 0x0) FAIL github.com/ethereum/go-ethereum/swarm/network/simulations/discovery 804.826s * swarm/network: run TestFileRetrieval with fewer nodes with -race Otherwise we get a failure due to extensive memory usage, as the CGO runtime cannot allocate more bytes. === RUN TestFileRetrieval ==7366==ERROR: ThreadSanitizer failed to allocate 0x80000 (524288) bytes of clock allocator (error code: 12) FATAL: ThreadSanitizer CHECK failed: ./gotsan.cc:6976 "((0 && "unable to mmap")) != (0)" (0x0, 0x0) FAIL github.com/ethereum/go-ethereum/swarm/network/stream 155.165s * swarm/network: run TestRetrieval with fewer nodes with -race Otherwise we get a failure due to extensive memory usage, as the CGO runtime cannot allocate more bytes ("ThreadSanitizer failed to allocate"). * swarm/network: skip flaky TestGetSubscriptionsRPC on Travis w/ -race Test fails a lot with something like: streamer_test.go:1332: Real subscriptions and expected amount don't match; real: 0, expected: 20 * swarm/storage: skip TestDB_SubscribePull* tests on Travis w/ -race Travis just hangs... ok github.com/ethereum/go-ethereum/swarm/storage/feed/lookup 1.307s keepalive keepalive keepalive or panics after a while. Without these tests the race detector job is now stable. Let's invetigate these tests in a separate issue: https://github.com/ethersphere/go-ethereum/issues/1245
2019-02-20 21:57:42 +00:00
if !testutil.RaceEnabled {
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
testSyncBetweenNodes(t, 4, dataChunkCount, true, 1)
testSyncBetweenNodes(t, 8, dataChunkCount, true, 1)
testSyncBetweenNodes(t, 16, dataChunkCount, true, 1)
}
2018-06-20 12:06:27 +00:00
}
func createMockStore(globalStore mock.GlobalStorer, id enode.ID, addr *network.BzzAddr) (lstore storage.ChunkStore, datadir string, err error) {
2018-06-20 12:06:27 +00:00
address := common.BytesToAddress(id.Bytes())
mockStore := globalStore.NewNodeStore(address)
params := storage.NewDefaultLocalStoreParams()
datadir, err = ioutil.TempDir("", "localMockStore-"+id.TerminalString())
2018-06-20 12:06:27 +00:00
if err != nil {
return nil, "", err
2018-06-20 12:06:27 +00:00
}
params.Init(datadir)
2018-06-20 12:06:27 +00:00
params.BaseKey = addr.Over()
lstore, err = storage.NewLocalStore(params, mockStore)
if err != nil {
return nil, "", err
}
return lstore, datadir, nil
2018-06-20 12:06:27 +00:00
}
func testSyncBetweenNodes(t *testing.T, nodes, chunkCount int, skipCheck bool, po uint8) {
sim := simulation.New(map[string]simulation.ServiceFunc{
"streamer": func(ctx *adapters.ServiceContext, bucket *sync.Map) (s node.Service, cleanup func(), err error) {
addr := network.NewAddr(ctx.Config.Node())
//hack to put addresses in same space
addr.OAddr[0] = byte(0)
netStore, delivery, clean, err := newNetStoreAndDeliveryWithBzzAddr(ctx, bucket, addr)
if err != nil {
return nil, nil, err
}
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
var dir string
var store *state.DBStore
Ci race detector handle failing tests (#19143) * swarm/storage: increase mget timeout in common_test.go TestDbStoreCorrect_1k sometimes timed out with -race on Travis. --- FAIL: TestDbStoreCorrect_1k (24.63s) common_test.go:194: testStore failed: timed out after 10s * swarm: remove unused vars from TestSnapshotSyncWithServer nodeCount and chunkCount is returned from setupSim and those values we use. * swarm: move race/norace helpers from stream to testutil As we will need to use the flag in other packages, too. * swarm: refactor TestSwarmNetwork case Extract long running test cases for better visibility. * swarm/network: skip TestSyncingViaGlobalSync with -race As panics on Travis. panic: runtime error: invalid memory address or nil pointer dereference [signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0x7e351b] * swarm: run TestSwarmNetwork with fewer nodes with -race As otherwise we always get test failure with `network_test.go:374: context deadline exceeded` even with raised `Timeout`. * swarm/network: run TestDeliveryFromNodes with fewer nodes with -race Test on Travis times out with 8 or more nodes if -race flag is present. * swarm/network: smaller node count for discovery tests with -race TestDiscoveryPersistenceSimulationSimAdapters failed on Travis with `-race` flag present. The failure was due to extensive memory usage, coming from the CGO runtime. Using a smaller node count resolves the issue. === RUN TestDiscoveryPersistenceSimulationSimAdapter ==7227==ERROR: ThreadSanitizer failed to allocate 0x80000 (524288) bytes of clock allocator (error code: 12) FATAL: ThreadSanitizer CHECK failed: ./gotsan.cc:6976 "((0 && "unable to mmap")) != (0)" (0x0, 0x0) FAIL github.com/ethereum/go-ethereum/swarm/network/simulations/discovery 804.826s * swarm/network: run TestFileRetrieval with fewer nodes with -race Otherwise we get a failure due to extensive memory usage, as the CGO runtime cannot allocate more bytes. === RUN TestFileRetrieval ==7366==ERROR: ThreadSanitizer failed to allocate 0x80000 (524288) bytes of clock allocator (error code: 12) FATAL: ThreadSanitizer CHECK failed: ./gotsan.cc:6976 "((0 && "unable to mmap")) != (0)" (0x0, 0x0) FAIL github.com/ethereum/go-ethereum/swarm/network/stream 155.165s * swarm/network: run TestRetrieval with fewer nodes with -race Otherwise we get a failure due to extensive memory usage, as the CGO runtime cannot allocate more bytes ("ThreadSanitizer failed to allocate"). * swarm/network: skip flaky TestGetSubscriptionsRPC on Travis w/ -race Test fails a lot with something like: streamer_test.go:1332: Real subscriptions and expected amount don't match; real: 0, expected: 20 * swarm/storage: skip TestDB_SubscribePull* tests on Travis w/ -race Travis just hangs... ok github.com/ethereum/go-ethereum/swarm/storage/feed/lookup 1.307s keepalive keepalive keepalive or panics after a while. Without these tests the race detector job is now stable. Let's invetigate these tests in a separate issue: https://github.com/ethersphere/go-ethereum/issues/1245
2019-02-20 21:57:42 +00:00
if testutil.RaceEnabled {
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
// Use on-disk DBStore to reduce memory consumption in race tests.
dir, err = ioutil.TempDir("", "swarm-stream-")
if err != nil {
return nil, nil, err
}
store, err = state.NewDBStore(dir)
if err != nil {
return nil, nil, err
}
} else {
store = state.NewInmemoryStore()
}
r := NewRegistry(addr.ID(), delivery, netStore, store, &RegistryOptions{
Retrieval: RetrievalDisabled,
Syncing: SyncingAutoSubscribe,
SkipCheck: skipCheck,
}, nil)
cleanup = func() {
r.Close()
clean()
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
if dir != "" {
os.RemoveAll(dir)
}
}
return r, cleanup, nil
},
})
defer sim.Close()
2018-06-20 12:06:27 +00:00
// create context for simulation run
timeout := 30 * time.Second
ctx, cancel := context.WithTimeout(context.Background(), timeout)
// defer cancel should come before defer simulation teardown
defer cancel()
_, err := sim.AddNodesAndConnectChain(nodes)
2018-06-20 12:06:27 +00:00
if err != nil {
t.Fatal(err)
2018-06-20 12:06:27 +00:00
}
result := sim.Run(ctx, func(ctx context.Context, sim *simulation.Simulation) (err error) {
nodeIDs := sim.UpNodeIDs()
2018-06-20 12:06:27 +00:00
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 22:59:00 +00:00
nodeIndex := make(map[enode.ID]int)
for i, id := range nodeIDs {
nodeIndex[id] = i
2018-06-20 12:06:27 +00:00
}
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
disconnected := watchDisconnections(ctx, sim)
defer func() {
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
if err != nil && disconnected.bool() {
err = errors.New("disconnect events received")
}
}()
2018-06-20 12:06:27 +00:00
// each node Subscribes to each other's swarmChunkServerStreamName
for j := 0; j < nodes-1; j++ {
id := nodeIDs[j]
client, err := sim.Net.GetNode(id).Client()
2018-06-20 12:06:27 +00:00
if err != nil {
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
return fmt.Errorf("node %s client: %v", id, err)
2018-06-20 12:06:27 +00:00
}
sid := nodeIDs[j+1]
client.CallContext(ctx, nil, "stream_subscribeStream", sid, NewStream("SYNC", FormatSyncBinKey(1), false), NewRange(0, 0), Top)
if err != nil {
return err
2018-06-20 12:06:27 +00:00
}
if j > 0 || nodes == 2 {
item, ok := sim.NodeItem(nodeIDs[j], bucketKeyFileStore)
if !ok {
return fmt.Errorf("No filestore")
}
fileStore := item.(*storage.FileStore)
size := chunkCount * chunkSize
_, wait, err := fileStore.Store(ctx, testutil.RandomReader(j, size), int64(size), false)
2018-06-20 12:06:27 +00:00
if err != nil {
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
return fmt.Errorf("fileStore.Store: %v", err)
2018-06-20 12:06:27 +00:00
}
wait(ctx)
2018-06-20 12:06:27 +00:00
}
}
// here we distribute chunks of a random file into stores 1...nodes
if _, err := sim.WaitTillHealthy(ctx); err != nil {
return err
2018-06-20 12:06:27 +00:00
}
// collect hashes in po 1 bin for each node
hashes := make([][]storage.Address, nodes)
totalHashes := 0
hashCounts := make([]int, nodes)
for i := nodes - 1; i >= 0; i-- {
if i < nodes-1 {
hashCounts[i] = hashCounts[i+1]
}
item, ok := sim.NodeItem(nodeIDs[i], bucketKeyDB)
if !ok {
return fmt.Errorf("No DB")
}
netStore := item.(*storage.NetStore)
netStore.Iterator(0, math.MaxUint64, po, func(addr storage.Address, index uint64) bool {
hashes[i] = append(hashes[i], addr)
totalHashes++
hashCounts[i]++
return true
})
2018-06-20 12:06:27 +00:00
}
var total, found int
for _, node := range nodeIDs {
i := nodeIndex[node]
for j := i; j < nodes; j++ {
total += len(hashes[j])
for _, key := range hashes[j] {
item, ok := sim.NodeItem(nodeIDs[j], bucketKeyDB)
if !ok {
return fmt.Errorf("No DB")
}
db := item.(*storage.NetStore)
_, err := db.Get(ctx, key)
if err == nil {
found++
}
2018-06-20 12:06:27 +00:00
}
}
log.Debug("sync check", "node", node, "index", i, "bin", po, "found", found, "total", total)
2018-06-20 12:06:27 +00:00
}
if total == found && total > 0 {
return nil
}
return fmt.Errorf("Total not equallying found: total is %d", total)
})
2018-06-20 12:06:27 +00:00
if result.Error != nil {
t.Fatal(result.Error)
2018-06-20 12:06:27 +00:00
}
}
//TestSameVersionID just checks that if the version is not changed,
//then streamer peers see each other
func TestSameVersionID(t *testing.T) {
//test version ID
v := uint(1)
sim := simulation.New(map[string]simulation.ServiceFunc{
"streamer": func(ctx *adapters.ServiceContext, bucket *sync.Map) (s node.Service, cleanup func(), err error) {
addr, netStore, delivery, clean, err := newNetStoreAndDelivery(ctx, bucket)
if err != nil {
return nil, nil, err
}
r := NewRegistry(addr.ID(), delivery, netStore, state.NewInmemoryStore(), &RegistryOptions{
Retrieval: RetrievalDisabled,
Syncing: SyncingAutoSubscribe,
}, nil)
bucket.Store(bucketKeyRegistry, r)
//assign to each node the same version ID
r.spec.Version = v
cleanup = func() {
r.Close()
clean()
}
return r, cleanup, nil
},
})
defer sim.Close()
//connect just two nodes
log.Info("Adding nodes to simulation")
_, err := sim.AddNodesAndConnectChain(2)
if err != nil {
t.Fatal(err)
}
log.Info("Starting simulation")
ctx := context.Background()
//make sure they have time to connect
time.Sleep(200 * time.Millisecond)
result := sim.Run(ctx, func(ctx context.Context, sim *simulation.Simulation) error {
//get the pivot node's filestore
nodes := sim.UpNodeIDs()
item, ok := sim.NodeItem(nodes[0], bucketKeyRegistry)
if !ok {
return fmt.Errorf("No filestore")
}
registry := item.(*Registry)
//the peers should connect, thus getting the peer should not return nil
if registry.getPeer(nodes[1]) == nil {
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
return errors.New("Expected the peer to not be nil, but it is")
}
return nil
})
if result.Error != nil {
t.Fatal(result.Error)
}
log.Info("Simulation ended")
}
//TestDifferentVersionID proves that if the streamer protocol version doesn't match,
//then the peers are not connected at streamer level
func TestDifferentVersionID(t *testing.T) {
//create a variable to hold the version ID
v := uint(0)
sim := simulation.New(map[string]simulation.ServiceFunc{
"streamer": func(ctx *adapters.ServiceContext, bucket *sync.Map) (s node.Service, cleanup func(), err error) {
addr, netStore, delivery, clean, err := newNetStoreAndDelivery(ctx, bucket)
if err != nil {
return nil, nil, err
}
r := NewRegistry(addr.ID(), delivery, netStore, state.NewInmemoryStore(), &RegistryOptions{
Retrieval: RetrievalDisabled,
Syncing: SyncingAutoSubscribe,
}, nil)
bucket.Store(bucketKeyRegistry, r)
//increase the version ID for each node
v++
r.spec.Version = v
cleanup = func() {
r.Close()
clean()
}
return r, cleanup, nil
},
})
defer sim.Close()
//connect the nodes
log.Info("Adding nodes to simulation")
_, err := sim.AddNodesAndConnectChain(2)
if err != nil {
t.Fatal(err)
}
log.Info("Starting simulation")
ctx := context.Background()
//make sure they have time to connect
time.Sleep(200 * time.Millisecond)
result := sim.Run(ctx, func(ctx context.Context, sim *simulation.Simulation) error {
//get the pivot node's filestore
nodes := sim.UpNodeIDs()
item, ok := sim.NodeItem(nodes[0], bucketKeyRegistry)
if !ok {
return fmt.Errorf("No filestore")
}
registry := item.(*Registry)
//getting the other peer should fail due to the different version numbers
if registry.getPeer(nodes[1]) != nil {
swarm: fix network/stream data races (#19051) * swarm/network/stream: newStreamerTester cleanup only if err is nil * swarm/network/stream: raise newStreamerTester waitForPeers timeout * swarm/network/stream: fix data races in GetPeerSubscriptions * swarm/storage: prevent data race on LDBStore.batchesC https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461775049 * swarm/network/stream: fix TestGetSubscriptionsRPC data race https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461768477 * swarm/network/stream: correctly use Simulation.Run callback https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-461783804 * swarm/network: protect addrCountC in Kademlia.AddrCountC function https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462273444 * p2p/simulations: fix a deadlock calling getRandomNode with lock https://github.com/ethersphere/go-ethereum/issues/1198#issuecomment-462317407 * swarm/network/stream: terminate disconnect goruotines in tests * swarm/network/stream: reduce memory consumption when testing data races * swarm/network/stream: add watchDisconnections helper function * swarm/network/stream: add concurrent counter for tests * swarm/network/stream: rename race/norace test files and use const * swarm/network/stream: remove watchSim and its panic * swarm/network/stream: pass context in watchDisconnections * swarm/network/stream: add concurrent safe bool for watchDisconnections * swarm/storage: fix LDBStore.batchesC data race by not closing it
2019-02-13 12:03:23 +00:00
return errors.New("Expected the peer to be nil, but it is not")
}
return nil
})
if result.Error != nil {
t.Fatal(result.Error)
}
log.Info("Simulation ended")
}