Commit Graph

360 Commits

Author SHA1 Message Date
Viktor Trón
0529015091
swarm/network: hive bug: needed shallow peers are not sent to nodes beyond connection's proximity order (#19326)
* swarm/network: fix hive bug not sending shallow peers

-  hive bug: needed shallow peers were not sent to nodes beyond connection's proximity order
- add extensive protocol exchange tests for initial subPeersMsg-peersMsg exchange
- modify bzzProtocolTester to allow pregenerated overlay addresses

* swarm/network: attempt to fix hive persistance test

* swarm/network: fix TestHiveStatePersistance (#1320)

* swarm/network: remove trace lines from the hive persistance test

* address PR review comments

* swarm/network: address PR comments on TestInitialPeersMsg

 * eliminate *testing.T argument from bzz/hive protocoltesters
 * add sorting (only runs in test code) on peersMsg payload
 * add random (0 to MaxPeersPerPO) peers for each po
 * add extra peers closer to pivot than control
2019-04-02 09:15:16 +02:00
lash
2f5b6cb442 swarm/network: Use different privatekey for bzz overlay in sim (#19313)
* cmd/swarm, p2p, swarm: Enable ENR in binary/execadapter

* cmd/p2p/swarm: Remove comments + config.Enode nomarshal

* p2p/simulations: Remove superfluous error check

* p2p/simulation: Move init enode comment

* swarm, p2p/simulations, cmd/swarm: Use nodekey in binary record sign

* swarm/network, swarm/pss: Dervice bzzkey

* swarm/pss: Remove unused function

* swarm/network: Store swarm private key in simulation bucket

* swarm/pss: Shorten TextProxNetwork shortrunning test timeout

* swarm/pss: Increase prox test timeout

* swarm/pss: Increase timeout slightly on shortrunning proxtest

* swarm/network: Simplify bucket instantiation in servicectx func

* p2p/simulations: Tcpport -> udpport

* swarm/network, swarm/pss: Simplify + correct lock in servicefunc sim

* swarm/network: Cleanup after rebase on extract swarm enode new

* p2p/simulations, swarm/network: Make exec disc test pass

* swarm/network: Prune ye olde comment

* swarm/pss: Correct revised bzzkey method call

* swarm/network: Clarify comment about privatekey generation data

* swarm/pss: Fix syntax errors after rebase

* swarm/network: Rename misleadingly named method

(amend commit to trigger ci - attempt 5)
2019-03-22 21:37:25 +01:00
gluk256
8d04154691 p2p/simulations: wait until all connections are recreated when uploading snapshot (#19312)
* swarm/network/simulation: test cases refactored

* swarm/pss: minor refactoring

* swarm/simulation: UploadSnapshot updated

* swarm/network: style fix

* swarm/pss: bugfix
2019-03-22 11:20:17 +01:00
lash
09924cbcaa cmd/swarm, p2p, swarm: Enable ENR in binary/execadapter (#19309)
* cmd/swarm, p2p, swarm: Enable ENR in binary/execadapter

* cmd/p2p/swarm: Remove comments + config.Enode nomarshal

* p2p/simulations: Remove superfluous error check

* p2p/simulation: Move init enode comment

* swarm/api: Check error in config test

* swarm, p2p/simulations, cmd/swarm: Use nodekey in binary record sign

* cmd/swarm: Make nodekey available for swarm api config
2019-03-22 05:55:47 +01:00
Anton Evangelatov
baded64d88
swarm/network: measure time of messages in priority queue (#19250) 2019-03-20 21:30:34 +01:00
gluk256
6e401792ce swarm/pss: negihbourhood addressing simulation tests (#19278)
* swarm/pss: fixed bug in pss.process, test added

* swarm/pss: test case updated

* swarm/pss: WaitTillSnapshotRecreated() func added

* swarm/pss: snapshot test updated

* swarm/pss: WaitTillSnapshotLoaded() fixed

* swarm/pss: gofmt applied

* swarm/pss: refactoring, file renamed

* swarm/pss: input data fixed

* swarm/pss: race condition fixed

* swarm/pss: test timeout increased

* swarm/pss: eliminated the global variables

* swarm/pss: tests added

* swarm/pss: comments added

* swarm/pss: comment fixed

* swarm/pss: refactored according to review

* swarm/pss: style fix

* swarm/pss: increased timeout
2019-03-16 08:39:38 +01:00
Janoš Guljaš
f180981273 swarm/shed: add vector uint64 field (#19279) 2019-03-15 15:12:45 +01:00
lash
4b4f03ca37 swarm, p2p: Prerequities for ENR replacing handshake (#19275)
* swarm/api, swarm/network, p2p/simulations: Prerequisites for handshake remove

* swarm, p2p: Add full sim node configs for protocoltester

* swarm/network: Make stream package pass tests

* swarm/network: Extract peer and addr types out of protocol file

* p2p, swarm: Make p2p/protocols tests pass + rename types.go

* swarm/network: Deactivate ExecAdapter test until binary ENR prep

* swarm/api: Remove comments

* swarm/network: Uncomment bootnode record load
2019-03-15 11:27:17 +01:00
Anton Evangelatov
1a3e25e4c1
swarm: tracing improvements (#19249) 2019-03-11 11:45:34 +01:00
Janoš Guljaš
9a58a9b91a swarm/storage/localstore: global batch write lock (#19245)
* swarm/storage/localstore: most basic database

* swarm/storage/localstore: fix typos and comments

* swarm/shed: add uint64 field Dec and DecInBatch methods

* swarm/storage/localstore: decrement size counter on ModeRemoval update

* swarm/storage/localstore: unexport modeAccess and modeRemoval

* swarm/storage/localstore: add WithRetrievalCompositeIndex

* swarm/storage/localstore: add TestModeSyncing

* swarm/storage/localstore: fix test name

* swarm/storage/localstore: add TestModeUpload

* swarm/storage/localstore: add TestModeRequest

* swarm/storage/localstore: add TestModeSynced

* swarm/storage/localstore: add TestModeAccess

* swarm/storage/localstore: add TestModeRemoval

* swarm/storage/localstore: add mock store option for chunk data

* swarm/storage/localstore: add TestDB_pullIndex

* swarm/storage/localstore: add TestDB_gcIndex

* swarm/storage/localstore: change how batches are written

* swarm/storage/localstore: add updateOnAccess function

* swarm/storage/localhost: add DB.gcSize

* swarm/storage/localstore: update comments

* swarm/storage/localstore: add BenchmarkNew

* swarm/storage/localstore: add retrieval tests benchmarks

* swarm/storage/localstore: accessors redesign

* swarm/storage/localstore: add semaphore for updateGC goroutine

* swarm/storage/localstore: implement basic garbage collection

* swarm/storage/localstore: optimize collectGarbage

* swarm/storage/localstore: add more garbage collection tests cases

* swarm/shed, swarm/storage/localstore: rename IndexItem to Item

* swarm/shed: add Index.CountFrom

* swarm/storage/localstore: persist gcSize

* swarm/storage/localstore: remove composite retrieval index

* swarm/shed: IterateWithPrefix and IterateWithPrefixFrom Index functions

* swarm/storage/localstore: writeGCSize function with leveldb batch

* swarm/storage/localstore: unexport modeSetRemove

* swarm/storage/localstore: update writeGCSizeWorker comment

* swarm/storage/localstore: add triggerGarbageCollection function

* swarm/storage/localstore: call writeGCSize on DB Close

* swarm/storage/localstore: additional comment in writeGCSizeWorker

* swarm/storage/localstore: add MetricsPrefix option

* swarm/storage/localstore: fix a typo

* swamr/shed: only one Index Iterate function

* swarm/storage/localstore: use shed Iterate function

* swarm/shed: pass a new byte slice copy to index decode functions

* swarm/storage/localstore: implement feed subscriptions

* swarm/storage/localstore: add more subscriptions tests

* swarm/storage/localsore: add parallel upload test

* swarm/storage/localstore: use storage.MaxPO in subscription tests

* swarm/storage/localstore: subscription of addresses instead chunks

* swarm/storage/localstore: lock item address in collectGarbage iterator

* swarm/storage/localstore: fix TestSubscribePull to include MaxPO

* swarm/storage/localstore: improve subscriptions

* swarm/storage/localstore: add TestDB_SubscribePull_sinceAndUntil test

* swarm/storage/localstore: adjust pull sync tests

* swarm/storage/localstore: remove writeGCSizeDelay and use literal

* swarm/storage/localstore: adjust subscriptions tests delays and comments

* swarm/storage/localstore: add godoc package overview

* swarm/storage/localstore: fix a typo

* swarm/storage/localstore: update package overview

* swarm/storage/localstore: remove repeated index change

* swarm/storage/localstore: rename ChunkInfo to ChunkDescriptor

* swarm/storage/localstore: add comment in collectGarbageWorker

* swarm/storage/localstore: replace atomics with mutexes for gcSize and tests

* swarm/storage/localstore: protect addrs map in pull subs tests

* swarm/storage/localstore: protect slices in push subs test

* swarm/storage/localstore: protect chunks in TestModePutUpload_parallel

* swarm/storage/localstore: fix a race in TestDB_updateGCSem defers

* swarm/storage/localstore: remove parallel flag from tests

* swarm/storage/localstore: fix a race in testDB_collectGarbageWorker

* swarm/storage/localstore: remove unused code

* swarm/storage/localstore: add more context to pull sub log messages

* swarm/storage/localstore: BenchmarkPutUpload and global lock option

* swarm/storage/localstore: pre-generate chunks in BenchmarkPutUpload

* swarm/storage/localstore: correct useGlobalLock in collectGarbage

* swarm/storage/localstore: fix typos and update comments

* swarm/storage/localstore: update writeGCSize comment

* swarm/storage/localstore: global batch write lock

* swarm/storage/localstore: remove global lock option

* swarm/storage/localstore: simplify DB.Close
2019-03-09 00:06:39 +01:00
Anton Evangelatov
bb55b0fb53 swarm/storage: add comparison towards leveldb.ErrNotFound (#19243)
* swarm/storage: add comparison towards leveldb.ErrNotFound

* swarm/storage: wrap leveldb ErrNotFound
2019-03-08 17:28:57 +01:00
Anton Evangelatov
2cfe0bed9f swarm: fix relationship between spans in open tracing (#19236)
* swarm/network: propagate span with ctx

* swarm/network: try to stop stream.send.request spans on time

* swarm/storage: add chunk ref as a log to netstore.fetcher span
2019-03-08 08:52:25 +01:00
Janoš Guljaš
a6e5c6a2cc swarm/storage/localstore: fix synchronization in TestDB_gcSize (#19235) 2019-03-08 05:59:59 +01:00
Janoš Guljaš
eb199f1fc2 swarm: localstore hasser (#19230) 2019-03-07 10:07:54 +01:00
Ferenc Szabo
d45f8d1880 swarm/network: remove *WithServer tests from stream package (#19223)
These tests never run as the build tag excluded them from the CI
execution. As a results the (dead) code got out of sync with other
parts of Swarm and now they would not even compile. => Removed.

resolves ethersphere/go-ethereum#1238
2019-03-07 09:27:56 +01:00
holisticode
a87776a5fe swarm/network/stream: Fix flaky tests in GetSubscriptionsRPC test (#19227)
* swarm/network/stream: fixed timing issues

* swarm/network/stream: only count first iteration of subscriptions

* swarm/network/stream/: fix linter errors
2019-03-07 09:24:28 +01:00
holisticode
81ed700157 Enable longrunning tests to run (#19208)
* p2p/simulations: increased snapshot load timeout for debugging

* swarm/network/stream: less nodes for snapshot longrunning tests

* swarm/network: fixed longrunning tests

* swarm/network/stream: store kademlia in bucket

* swarm/network/stream: disabled healthy check in delivery tests

* swarm/network/stream: longer SyncUpdateDelay for longrunning tests

* swarm/network/stream: more debug output

* swarm/network/stream: reduced longrunning snapshot tests to 64 nodes

* swarm/network/stream: don't WaitTillHealthy in SyncerSimulation

* swarm/network/stream: cleanup for PR
2019-03-05 12:54:46 +01:00
Janoš Guljaš
216bd2ceba swarm/storage/localstore: fix testDB_collectGarbageWorker data race (#19206) 2019-03-04 22:19:57 +01:00
Anton Evangelatov
f9aa1cd21f Revert "swarm/network: Use actual remote peer ip in underlay (#19137)" (#19193)
This reverts commit 460d206f30.
2019-03-02 08:45:07 +01:00
Janoš Guljaš
b797dd07d2 swarm/shed, swarm/storage/localstore: add LastPullSubscriptionChunk (#19190)
* swarm/shed, swarm/storage/localstore: add LastPullSubscriptionChunk

* swarm/shed: fix comments

* swarm/shed: fix TestIncByteSlice test

* swarm/storage/localstore: fix TestDB_LastPullSubscriptionChunk
2019-03-02 08:44:22 +01:00
Anton Evangelatov
4e9230ea7a
swarm: enable p2p/discovery and disable dynamic dialling (#19189) 2019-03-01 12:20:37 +01:00
holisticode
994326ba00 swarm: new snapshot files (#19185) 2019-02-28 22:30:36 +01:00
lash
62d9d63858 swarm/network: WIP consider all nodes for healthy iteration (#19155)
* swarm/network: WIP consider all nodes for healthy iteration

* swarm/network/simulation: extend TestWaitTillHealthy to really check kads are healthy

* cmd/swarm/swarm-snapshot: fixed bugs in snapshot creation binary

* swarm/network/simulation: addressed PR comments

* swarm/network/simulation: defer sim.Clsoe()

* swarm/network/simulation: fixed wrong sim.Close()

* swarm/network/simulation: addressed PR comments

* cmd/swarm/swarm-snapshot: reducing default to 8 nodes, more to 4

* cmd/swarm/swarm-snapshot: extended timeout to 3 mins, or 256 nodes snapshot times out

* swarm/network/simulation: More PR comments
2019-02-28 08:12:50 +01:00
Janoš Guljaš
f0233948d2 swarm/chunk: move chunk related declarations to chunk package (#19170) 2019-02-26 16:09:32 +01:00
Janoš Guljaš
c83ba9e794 swarm/storage/localstore: fix tests for windows os (#19161) 2019-02-26 10:02:05 +01:00
Janoš Guljaš
340a53a98b swarm/pss: fix data race on HandshakeController.symKeyIndex (#19162)
* swarm/pss: fix data race on HandshakeController.symKeyIndex

The HandshakeController.symKeyIndex map was accessed concurrently.
Since insufficient test coverage the race is not detected every time.
However, running TestClientHandshake a 100 times seems to be enough to
reproduce the race.

Note: I've chosen HandshakeController.lock to protect
HandshakeController.symKeyIndex as that was already protected in a few
functions by that lock.

Additionally:
- removed unused testStore
- enabled tests in handshake_test.go as they pass
- removed code duplication by adding getSymKey()

* swarm/pss: fix a data race on HandshakeController.keyC

* swarm/pss: fix data races with on Pss.symKeyPool
2019-02-26 08:17:20 +01:00
Janoš Guljaš
872370e3bc swarm/network/simulation: do not copy node mutex in UploadSnapshot (#19160) 2019-02-25 10:03:31 +01:00
Matthew Halpern
81babe1509 swarm/*: remove redundant type specifiers (#19089) 2019-02-25 08:58:18 +01:00
Matthew Halpern
4e87b444ca swarm/pss: remove unused function (#19100) 2019-02-24 12:41:22 +01:00
Matthew Halpern
90b6cdaadf cmd,swarm: enforce camel case variable names (#19060) 2019-02-24 12:39:23 +01:00
Janoš Guljaš
64d10c0872 swarm: mock store listings (#19157)
* swarm/storage/mock: implement listings methods for mem and rpc stores

* swarm/storage/mock/rpc: add comments and newTestStore helper function

* swarm/storage/mock/mem: add missing comments

* swarm/storage/mock: add comments to new types and constants

* swarm/storage/mock/db: implement listings for mock/db global store

* swarm/storage/mock/test: add comments for MockStoreListings

* swarm/storage/mock/explorer: initial implementation

* cmd/swarm/global-store: add chunk explorer

* cmd/swarm/global-store: add chunk explorer tests

* swarm/storage/mock/explorer: add tests

* swarm/storage/mock/explorer: add swagger api definition

* swarm/storage/mock/explorer: not-zero test values for invalid addr and key

* swarm/storage/mock/explorer: test wildcard cors origin

* swarm/storage/mock/db: renames based on Fabio's suggestions

* swarm/storage/mock/explorer: add more comments to testHandler function

* cmd/swarm/global-store: terminate subprocess with Kill in tests
2019-02-23 10:47:33 +01:00
Janoš Guljaš
02c28046a0 swarm: Fix localstore test deadlock with race detector (#19153)
* swarm/storage/localstore: close localstore in two tests

* swarm/storage/localstore: fix a possible deadlock in tests

* swarm/storage/localstore: re-enable pull subs tests for travis race

* swarm/storage/localstore: stop sending to errChan on context done in tests

* swarm/storage/localstore: better want check in readPullSubscriptionBin

* swarm/storage/localstore: protect chunk put with addr lock in tests

* swamr/storage/localstore: wait for gc and writeGCSize workers on Close

* swarm/storage/localstore: more correct testDB_collectGarbageWorker

* swarm/storage/localstore: set DB Close timeout to 5s
2019-02-22 23:19:09 +01:00
Janoš Guljaš
c8da76e63d swarm/shed: fix a deadlock in meter function (#19149) 2019-02-21 20:42:53 +01:00
Janoš Guljaš
836c846812 swarm/network/master: protect SetNextBatch iterator after close (#19147) 2019-02-21 18:33:49 +01:00
Péter Szilágyi
b9808e392f
swarm/version: bump to v0.3.12 unstable 2019-02-21 15:25:42 +02:00
Matthew Halpern
fbedf62f3d swarm/storage: fix loop bound for database cleanup (#19085)
The current loop continuation condition is always true as a uint8
is always being checked whether it is less than 255 (its maximum
value). Since the loop starts with the value 1, the loop termination
can be guarranteed to exit once the value overflows to 0.
2019-02-21 06:37:32 +01:00
Ferenc Szabo
e38b227ce6 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 22:57:42 +01:00
lash
d36e974ba3 swarm/network: Keep span across roundtrip (#19140)
* swarm/newtork: WIP Span request span until delivery and put

* swarm/storage: Introduce new trace across single fetcher lifespan

* swarm/network: Put span ids for sendpriority in context value

* swarm: Add global span store in tracing

* swarm/tracing: Add context key constants

* swarm/tracing: Add comments

* swarm/storage: Remove redundant fix for filestore

* swarm/tracing: Elaborate constants comments

* swarm/network, swarm/storage, swarm:tracing: Minor cleanup
2019-02-20 14:50:37 +01:00
lash
460d206f30 swarm/network: Use actual remote peer ip in underlay (#19137)
* swarm/network: Logline to see handshake addr

* swarm/network: Replace remote ip in handshake uaddr

* swarm/network: Add test for enode uaddr rewrite method

* swarm/network: Remove redundance pointer return from sanitize

* swarm/network: Obeying the linting machine

* swarm/network: Add panic comment

(travis trigger take 1)
2019-02-20 14:46:00 +01:00
Janoš Guljaš
ba2dfa5ce4 swarm/network/stream: fix a goroutine leak in Registry (#19139)
* swarm/network/stream: fix a goroutine leak in Registry

* swarm/network, swamr/network/stream: Kademlia close addr count and depth change chans

* swarm/network/stream: rename close channel to quit

* swarm/network/stream: fix sync between NewRegistry goroutine and Close method
2019-02-20 14:45:25 +01:00
lash
d88c6ce6b0 swarm: Reinstate Pss Protocol add call through swarm service (#19117)
* swarm: Reinstate Pss Protocol add call through swarm service

* swarm: Even less self
2019-02-18 16:44:50 +01:00
Ferenc Szabo
50b872bf05 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 85a79b3ad3,
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 07:38:14 +01:00
gluk256
12ca3b172a swarm/pss: refactoring (#19110)
* swarm/pss: split pss and keystore

* swarm/pss: moved whisper to keystore

* swarm/pss: goimports fixed
2019-02-17 06:29:41 +01:00
Elad
5b8ae7885e swarm/storage: fix influxdb gc metrics report (#19102) 2019-02-15 07:41:42 +01:00
holisticode
2af24724dd swarm/network: Saturation check for healthy networks (#19071)
* swarm/network: new saturation for  implementation

* swarm/network: re-added saturation func in Kademlia as it is used elsewhere

* swarm/network: saturation with higher MinBinSize

* swarm/network: PeersPerBin with depth check

* swarm/network: edited tests to pass new saturated check

* swarm/network: minor fix saturated check

* swarm/network/simulations/discovery: fixed renamed RPC call

* swarm/network: renamed to isSaturated and returns bool

* swarm/network: early depth check
2019-02-14 19:01:50 +01:00
Elad
3ee09ba035 swarm/storage/netstore: add fetcher cancellation on shutdown (#19049)
swarm/network/stream: remove netstore internal wg
swarm/network/stream: run individual tests with t.Run
2019-02-14 07:51:57 +01:00
Janoš Guljaš
3fd6db2bf6 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 13:03:23 +01:00
Elad
d596bea2d5 swarm: fix uptime gauge update goroutine leak by introducing cleanup functions (#19040) 2019-02-13 08:15:03 +01:00
holisticode
3d22a46c94 swarm/storage: fix HashExplore concurrency bug ethersphere#1211 (#19028)
* swarm/storage: fix HashExplore concurrency bug ethersphere#1211

*  swarm/storage: lock as value not pointer

* swarm/storage: wait for  to complete

* swarm/storage: fix linter problems

* swarm/storage: append to nil slice
2019-02-13 00:17:44 +01:00
gluk256
b30109df3c swarm/pss: mutex lifecycle fixed (#19045) 2019-02-13 00:12:41 +01:00