c25f8d4adf
sqlite kvstore backend ( #749 )
2020-02-20 11:49:34 +01:00
9e58e22761
[skip ci] remove an obsolete comment
2020-02-20 10:58:38 +02:00
5742299c94
Fix freezes in connection initialization.
2020-02-20 10:58:38 +02:00
880b981a0a
Don't wait for 10 seconds when connecting to the network
2020-02-19 12:45:34 +02:00
55886ff0ac
More proper fix replacing hotfix 052a6ac3bc
...
nim-eth was bumped to take advantage of a fixed `==` for public eth keys
2020-02-19 11:59:38 +02:00
3b9e9fefa8
Make the persistent network keys compatible with the Go daemon
...
The daemon requires its key file to be encoded as ProtoBuf value
with a proper `key scheme` discriminator.
2020-02-19 10:58:10 +02:00
1b1cd915db
Support ENR boot records with the Go daemon
2020-02-19 10:26:32 +02:00
052a6ac3bc
Hotfix: disable the bootstrap node connectivity check
...
This used to behave properly before the rebase, but currently
it forces the bootstrap node to exit, because it ends up being
launched with an ENR list telling it to connect to itself.
The root cause will be investigated in a follow-up PR.
2020-02-18 20:42:25 +02:00
57811dce89
Don't log unwanted lines in beacon_node --help
2020-02-18 12:53:49 +02:00
b2d5aba749
Advertise the TCP port of the bootstrap node properly
2020-02-18 12:53:49 +02:00
2a3e40e298
Fix test_discovery_helpers
2020-02-18 12:53:49 +02:00
2c814db750
Temporary switch to supporting only ENR bootstrap addresses
...
Turns out the DiscV5 code relies heavily on the presence of ENR
records at the moment, so we cannot drive it with ENodes. @kdeme
is working on refactoring that will relax these requirements.
2020-02-18 12:53:49 +02:00
7f395e554a
WIP Working local network sim with discovery
...
* The bootstrap_nodes.txt file in the node's data dir is now optional
* Log more data on start-up
* Use the latest ENR APIs
* Fix simulation build errors
2020-02-18 12:53:49 +02:00
d2f59ebab9
Switch to ENR records in the bootstrap_nodes.txt file
2020-02-18 12:53:49 +02:00
ee7c2c9dff
Unify the bootstrap nodes handling code
...
We no longer discriminate between ENR, MultiAddress or ENode
bootstrap records (all of them are remapped to ENodes).
The discovery loop will stochastically try to reconnect to
accidentally disconnected nodes.
2020-02-18 12:53:49 +02:00
4fb654f2af
Remove some SSZ compilation warnings
2020-02-18 12:53:49 +02:00
a620c83916
Handle status bar spanning multiple lines (relevant for the Tmux network sim)
2020-02-18 12:53:49 +02:00
bde1eefdcf
improve option naming
2020-02-17 22:00:41 +00:00
e13846bec6
allow beacon_node to verify finalization when appropriate, and enable by default for make eth2_network_simulation
2020-02-17 22:00:41 +00:00
d99ce1bcf0
fix json compile
2020-02-08 19:42:45 +01:00
521b0ed6ba
Clean 20200205 ( #729 )
...
* beacon node code cleanup
* rudimentary error checking on mainnet monitor
* start client even when sending deposit
* work around missing block number exception
* connect to testnet with web3 url
* pretty-print digests in json
2020-02-07 08:13:38 +01:00
09d735212d
initial refactoring of block sanity test runner, with several new tests ( #736 )
...
* initial refactoring of block sanity test runner, with several new tests enabled
* remove trailing whitespace
2020-02-07 08:11:26 +01:00
7efd113a75
store empty slot states also
2020-02-05 19:24:07 +00:00
fb9c4fabf4
fix state rewind
...
* rewind fast path comparison was not taking skipped slots into account
properly
* less messy blockref creation
2020-02-05 19:24:07 +00:00
b39f36b49b
use floodsub instead of gossipsub
2020-02-05 17:20:04 +01:00
1ffc2df23d
add a couple new deposit tests; fix the false-positive BLS verifications while keeping all but two tests working, despite mismatched 0.9/0.10 BLS standards; better-factor the skipping of BLS validation and Merkle tree validation
2020-02-04 18:39:38 +00:00
e4922cb177
implement compute_signing_root() and mark some functions as 0.10.1-compatible
2020-02-04 12:24:30 +00:00
45dd12cf3f
update process_deposit() to actually check is_valid_merkle_branch() unless skipValidation specified
2020-01-30 09:31:08 +00:00
98dc701473
Add PeerPool.addPeer async version and tests.
2020-01-29 15:28:41 +00:00
db20fc1172
Fix SyncQueue push(data) bug.
...
Rename lastSlot to HeadSlot.
Add failure test.
2020-01-29 15:28:41 +00:00
73dc72583f
Initial commit.
2020-01-29 15:28:41 +00:00
1f1a244f3f
Implement more checks and wire up the invalid SSZ tests
2020-01-29 10:10:28 +02:00
182593ccb5
Tenative fix for the test suite
2020-01-29 10:10:28 +02:00
2a3c237bbb
Handle malformed SSZ inputs properly
2020-01-29 10:10:28 +02:00
60b1775879
update check_attestation/process_attestation() to 0.10.1 and fill in missing check
2020-01-28 12:07:33 +02:00
3efbe36c30
additional 0.10.1 updates where already-implemented code is already done
2020-01-27 12:50:54 +00:00
dab107fa1e
more 0.10.1 updates
2020-01-27 12:50:54 +00:00
1ba2cebd85
v0.10.1 (test vectors + unchanged parts) ( #696 )
...
* update to 0.10.1
* SSZ Generic and nbench uses the v0.10.1 fixtures
* Tests + spec links: v0.10.0 -> v0.10.1
* Add v0.10.1 TODO in get_latest_attesting_balance (forkchoice)
* SSZ Bytes are now ByteList
* Remove nim-result submodules that was leftover/added by mistake in the branch
2020-01-27 10:56:32 +00:00
270c25c8b8
drop rlpx support ( #679 )
2020-01-24 09:32:52 +01:00
7dc2434e0d
clean up attestation logging
2020-01-23 18:48:26 +01:00
9cade2a5c4
LMDB: fix comment
2020-01-23 18:23:27 +02:00
d9c98b15c8
fix block pool finalization point init
2020-01-23 14:34:58 +02:00
95437e103a
disable state pruning
...
* fix crash when state root is present but state is missing
* fix state root removal when state is removed
* fix block pool initialization which needs tail state
* remove tail block pruning
* incomplete - fork states are not pruned
* incomplete - fork blocks are not pruned
* incomplete - empty slot states are not pruned
* unknown - tail/finalized block on empty slot might be incorrect
2020-01-22 19:45:38 +02:00
50f633209c
mark a couple dozen 0.9.2/0.9.4 spec locations as still consistent with 0.10.0; no functional changes
2020-01-22 13:03:45 +00:00
6cfa28e1f5
fix finalization cleanup
...
* add block pool finalization test
2020-01-21 15:53:53 +02:00
0743b3507c
Fix the build after the rebase
2020-01-21 01:26:57 +02:00
63b98f0334
Less confusing debug logging
2020-01-21 01:26:57 +02:00
c65575d105
Improved logging related to bootstrap nodes
2020-01-21 01:26:57 +02:00
589a3f06df
Fix an error introduced during rebasing
2020-01-21 01:26:57 +02:00
4848b12050
Detect and use Lighthouse's boot_enr.yaml files
2020-01-21 01:26:57 +02:00
07d329acb8
Wire-up the peer dialing loop and introduce --max-peers option
2020-01-21 01:26:57 +02:00
52878405b7
Integrate Discovery V5 and support ENR bootstrap records
2020-01-21 01:26:57 +02:00
74fc34eef2
Restore some constants to pass the constants & presets test
2020-01-21 01:26:57 +02:00
7e356b3fa6
Restore compilation with trace-level logging; Fixes #642
2020-01-21 01:26:57 +02:00
095b271bd4
Implement a loader for custom const presets
...
The loader has been tested with the presets published by Lighthouse.
You can try connecting to one of their testnets by running:
cd nim-beacon-chain
./connect-to-testnet lighthouse/testnet0
2020-01-21 01:26:57 +02:00
7a8054d36d
clean up block pool
...
* remove BlockPool.blocksBySlot (unused)
* simplify head pruning condition
* add head list smoke tests
* additional logging
2020-01-21 00:54:23 +02:00
14e1e3af52
LMDB: create db dir
2020-01-20 13:39:37 +00:00
7113259137
lmdb: 32-bit Windows fixes
2020-01-20 13:39:37 +00:00
2a67ac3c05
trie -> kv store
...
* simplify data storage to key-value, tries are not relevant for NBC
* locked-down version of lmdb dependency
* easier to build / maintain on various platforms
2020-01-20 13:39:37 +00:00
d853d7c024
remove SigKey.init (unused, fixes #664 )
2020-01-18 10:22:17 +01:00
3439763a9e
fix beacon node init
...
* update forkversion correctly
* ensure supplied state snapshot starts at genesis - we don't support
anything else
* more database sanity checks
2020-01-17 21:26:28 +00:00
c301e5d42f
mark 49 0.9.2/0.9.4 spec functions/data types as 0.10.0
2020-01-15 12:34:16 +00:00
860be026e1
fix block pool init head selection
...
the head state is not necessarily saved in the database, so we need to
make sure we update things to the correct place
2020-01-15 12:17:49 +00:00
b49003988a
Add Epoch Processing to nbench
2020-01-15 10:03:56 +00:00
8eaebf7163
update to spec version 0.10.0
2020-01-14 18:17:16 +00:00
2f68ee0ffe
new libp2p API
2020-01-13 20:26:19 +01:00
5d1e20012a
import base58 from stew
2020-01-13 20:02:29 +01:00
7c4d636f83
undo try blocks; that's not the traceback's source
2020-01-13 19:02:38 +01:00
468e0815b3
not a CatchableError
2020-01-13 18:53:59 +01:00
003a4c90be
sendErrorResponse(): don't crash when unable to send to peer
2020-01-13 17:12:49 +01:00
148527c716
slot fixes
...
* fix slot time navigation, add tests
* skip block proposal if head is more recent already - shouldn't happen
* use correct head when attesting to previous blocks
* log slot start/end processing
2019-12-23 17:56:56 +00:00
46de726e3d
mark all compatible 0.9.3 functions and data structures as 0.9.4
2019-12-20 17:24:42 +00:00
106352aff3
Nbench - Flexible benchmarking of Nimbus internals ( #641 )
...
* nbench PoC
* Remove the yaml files from the example scenarios
* update README with current status
* Add an alternative implementation that uses defer
* Forgot to add the old proc body
* slots-processing
* allow benching state_transition failures
* Add Attestations processing (workaround confutils bug:
- https://github.com/status-im/nim-confutils/issues/10
- https://github.com/status-im/nim-confutils/issues/11
- https://github.com/status-im/nim-confutils/issues/12
* Add CLI command in the readme
* Filter report and add notes about CPU cycles
* Report averages
* Add debugecho style time/cycle print
* Report when we skip BLS and state root verification
* Update to 0.9.3
* Generalize scenario parsing
* Support all block processing scenarios
* parallel bench runner PoC
* gitBetter load issues reporting (the load issues were invalid signature and expected to fail)
2019-12-20 17:14:43 +01:00
417f96213d
fix genesis interop by fixing deposit signature handling; add TRACE-level compilation to CI
2019-12-20 15:26:48 +00:00
746659bdc6
bump stew, fix warnings ( #655 )
2019-12-20 14:25:33 +01:00
b994da78a7
Move some attestation/block logic out of beacon node
...
* state data cache in block pool
* keep head state around
* more attestation logic in attestation pool
* first fork choice tests (!)
* fix fork choice (it's still likely broken / out of date)
2019-12-19 16:17:45 +00:00
ea4afd7454
mark 19 more 0.9.2-compatible entities as 0.9.3-compatible
2019-12-19 11:46:30 +00:00
350a6c071a
mark a couple compatible dozen 0.9.2 functions/data structures as 0.9.3
2019-12-18 16:13:39 +00:00
0328cd49bf
mark several more functions and data structures as 0.9.3
2019-12-18 10:00:17 +00:00
1610981be6
mark several more spec-referencing code/data parts as 0.9.3
2019-12-17 16:51:46 +00:00
27bf8908a9
mark a couple dozen functions and data structures as 0.9.3; fix TRACE-level compiling; stop downloading 0.9.2 test vectors; update cosmetic/display spec versions in test suite
2019-12-17 16:51:46 +00:00
22c44d254c
fix decoding for 0.9.3 types
2019-12-17 11:25:36 +01:00
15f289a281
Temporary switch back to the LibP2P daemon as the default choice
2019-12-17 00:13:51 +02:00
c824416f56
initial 0.9.3 spec update
2019-12-17 00:06:03 +02:00
8c104a0b94
inspector: add readme entry, decoding option
2019-12-16 16:08:30 +02:00
963f23e5ea
remove obsolete insecure comment
2019-12-16 13:36:39 +01:00
c66ca39cee
update heads when adding resolved blocks
2019-12-16 12:18:08 +02:00
4e85b50cb6
update head while syncing
2019-12-16 12:18:08 +02:00
741570113b
properly construct attestation data for selected slot
2019-12-13 09:04:30 +00:00
fbdb078eb3
fix return type to properly construct AggregateAndProof; check ATTESTATION_PROPAGATION_SLOT_RANGE as required, maintaining consistent slot offset
2019-12-13 09:04:30 +00:00
5880fc22ad
roadmap rest of naive attestation aggregation
2019-12-13 09:04:30 +00:00
c4daa726b5
fix AggregateAndProof definition
2019-12-13 09:04:30 +00:00
e2c68e075f
outline general flow of naive attestation aggregation
2019-12-13 09:04:30 +00:00
294db6aaca
initially implement specified attestation aggregation functions and data structures
2019-12-13 09:04:30 +00:00
cb1bc6cbf1
move attesting up from halfway to one third of the way through slots ( #643 )
...
* per honest validator and naïve/simple aggregator attestation specs, move attesting up from halfway to one third of the way through slots
* Update beacon_chain/beacon_node.nim
Co-Authored-By: Jacek Sieka <jacek@status.im>
2019-12-11 14:02:07 +01:00
a3145f0f5c
cleanups
...
* move stuff out of beaconnode
* fix use-after-free in query
2019-12-10 14:47:52 +00:00
10e6d4842f
Local sim finalizes with native LibP2P
2019-12-10 15:02:16 +02:00
98656377a3
Eliminate the code duplication in the LibP2P back-ends
2019-12-10 15:02:16 +02:00
889031453f
The beacon node can be compiled with the LibP2P native back-end
...
This hasn't been tested yet even in local sim.
2019-12-10 15:02:16 +02:00
eea523a89e
logging fixes
...
* don't log out-of-sync when there's no work to do
* small logging fixes
2019-12-10 10:18:30 +00:00
8ab0248209
Revert "tighten a few imports"
...
This reverts commit b24a9200d9
2019-12-10 02:18:47 +02:00
b24a9200d9
tighten a few imports
2019-12-09 12:58:28 +00:00
eef040954c
remove unused hasSigningRoot(...); stop running time-consuming-but-stubbed-out tests; add timing for peer connection tests
2019-12-06 16:48:32 +00:00
362ef752dc
dedeplicate some functions duplicated between libp2p_backend and libp2p_daemon_backend; remove some unused declared variables
2019-12-06 12:55:13 +00:00
5729ceb58e
rename notice field
2019-12-06 12:12:59 +00:00
86f267bc5d
switch debug to notice and add some contextual information
2019-12-06 12:12:59 +00:00
c762c0232d
render get_beacon_proposer_index(...) return value an option and remove overly aggressive assertion formerly causing crashes when missing validators ( #617 )
...
* render get_beacon_proposer_index(...) return value an option and remove overly aggressive assertion formerly causing crashes when missing validators
* follow addBlock(...) after refactoring
* [skip ci] fix typo
2019-12-04 11:49:59 +01:00
3b3a2b10f1
First attempt to have some fuzz tests exported for the fuzzing framework
2019-12-03 14:33:47 +02:00
a1198e5532
web3 fixes suggested by @yglukhov [skip ci]
2019-12-03 14:10:47 +02:00
afc0686b36
save ssz dummps during processing ( #471 )
2019-12-03 11:32:27 +00:00
c457904a61
avoid recursion when getting lmd ancestors
2019-12-03 13:24:16 +02:00
a641331448
Reduce the sync-related tracing in the testnets [skip ci]
2019-12-03 13:07:43 +02:00
ee2448b221
Simple validator onboarding
...
On your very first connection to each testnet, you'll be asked to
become a validator. Please consult our private repo for a Goerli
Eth1 private key that you can use for deposits.
Other changes:
* Added a simple wrapper ./connect-to-testnet script calling the
nims file in the correct environment. No extension was used to
make the command the same on Unix and Windows.
* Bumped a number of modules with fixes from this week
* `make testnet0` and `make testnet1` will no longer delete your
existing database. This is considered a more appropriate behavior
for testing forward sync.
2019-12-03 02:23:17 +02:00
7682a010a9
remove getCurrentExceptionMsg()
2019-12-02 22:06:42 +02:00
e88e559caf
render visible gradual validator emptying on catchup
2019-12-02 18:24:30 +02:00
90212eed2a
avoid producing blocks/attestations when out of sync
...
this is a temporary measure until we figure something better out - as it
stands, we'll advance with empty slots and crash because all validators
are out.
2019-12-02 15:26:42 +00:00
2a728d9468
fix stack overflow in isAncestor recursion
2019-11-28 17:36:07 +02:00
c98a0d28ae
add workaround for stack overflow on block add
2019-11-28 17:36:07 +02:00
8f2e523d26
remove a few hints and warnings ( #603 )
2019-11-28 13:50:19 +01:00
63f49eff76
document slot better
2019-11-27 15:14:20 +01:00
e6ae9977b2
clarify slot time
2019-11-27 14:59:54 +01:00
8c3ea1cabf
Handle missing blocks better in block_pool.getBlockRange
2019-11-27 13:45:34 +02:00
897176761d
even more tracing for testnet1 [skip ci]
2019-11-27 11:21:45 +02:00
6f650ca144
avoid pointless stream usage for main basic type in SSZ objects ( #596 )
...
* avoid pointless stream usage for uint64 in SSZ objects
2019-11-26 18:22:36 +00:00
fa22ba22b9
PeerPool initial commit.
2019-11-26 17:16:55 +00:00
1099548775
Restore the number of simulation nodes to 4; More block pool tracing [skip ci]
2019-11-26 19:02:56 +02:00
bde20436e8
~50% speedup by fast-pathing Eth2Digest merkle hashing
2019-11-26 15:23:27 +00:00
843c461af6
Remove some debugging output
2019-11-26 02:47:41 +02:00
8b89bbe391
Fix off-by-one error in syncing
2019-11-26 02:45:25 +02:00
1d8bfd8c0d
Allow overriding the bootstrap node in run_node.sh
2019-11-26 02:45:25 +02:00
8de4d38e4e
More tracing
2019-11-26 02:45:25 +02:00
ff3da4e152
more network traffic tracing
2019-11-26 02:45:25 +02:00
957645d752
Fix the bulds with json logging enabled
2019-11-26 02:45:25 +02:00
acd490916f
Attempt to fix the failing tests in CI
2019-11-26 02:45:25 +02:00
258f6f6334
detailed tracing in block_pool.getBlockRange
2019-11-26 02:45:25 +02:00
4c2ded25a6
Add detailed tracing for blocks by range requests
2019-11-26 02:45:25 +02:00
777b3f4e29
State/block pruning
2019-11-26 02:45:25 +02:00
76d3e74b02
fix head update message
2019-11-25 18:50:45 +01:00
a82869c03d
add validators even if they're not in startup state
2019-11-25 18:31:22 +02:00
1494bcc262
replace links to Apache and MIT licenses with HTTPS versions ( #592 )
2019-11-25 16:30:02 +01:00
904283e33e
make mount(...) public again
2019-11-25 14:48:59 +00:00
b82328b148
convert hash_tree_roots_prefix(...) to an iterator
2019-11-25 14:48:59 +00:00
c3d2634b97
re-export beginRecord(...)
2019-11-25 14:48:59 +00:00
f6de87b80e
remove O(n^2) eth1 deposit processing and scaffolding comparing it with O(n) deposit processing; re-enable now-viable-to-run genesis state initialization tests with around 16,384 validators
2019-11-25 14:48:59 +00:00
fa2a703a17
switch more funcs and procs from global to module scope; create scaffolding for non-O(n^2) eth1 deposit processing with assertions for equivalent functionality; fix a few more shellcheck warnings
2019-11-25 14:48:59 +00:00
fd4de5de0f
start untangling beaconnode ( #589 )
...
* Move BeaconNode type to its own file (fewer imports)
* disentangle sync protocol/request manager
* fix some old nimisms
* de-fear some logs
* simplify eth1 data production
* add stack tracing to release builds
* drop release compile flag for testnet
2019-11-25 15:36:25 +01:00
2e875ea17e
Better connection error handling in MainchainMonitor
2019-11-25 13:45:53 +02:00
a8de94ad05
add validator_leaving logging at debug chronicles level for all call paths to initiate_validator_exit(...)
2019-11-25 09:08:15 +00:00
8d9fdb0afd
miscellanous cleanups and removing a few funcs/procs from unnecessary global scopes
2019-11-25 09:08:15 +00:00
1938379bcd
Automated beacon constant checks ( #583 )
...
* Fix SSZ bitlist deserialization super silly bug
* Add an automated sanity checks of the beacon chain constants
* Remove SSZ consensus skipping procs [skip ci]
* Add phase 1 domains
* Fix mainnet constants
* Add missing phase 1 constants on minimal (they are not needed somehow on mainnet)
* Rebase artifact: constants were defined twice
2019-11-22 20:56:39 +01:00
Jacek Sieka
Zahary Karadjov
cheatfate
Dustin Brody
Dmitriy Ryajov
Mamy Ratsimbazafy
Ștefan Talpalaru
kdeme
Yuriy Glukhov