* Jenkins: add timeouts to testnet finalisation tests
- add nicer child process management to "launch_local_testnet.sh"
- correct a comment in "start.sh"
* the Jenkins worker's version of "timeout" doesn't support "-v"
* fix mainnet finalization and swith eth2_network_simulation to a kind of small-mainnet profile
* Fix slot reference in trace logging
* bump a couple of spec refs from v0.11.0 to v0.11.1
* bump another spec ref to v0.11.1, one more try at Jenkins test vector download CI issue
* fix other slot reference in trace logging and skip past single-block/multi-slot gaps to re-approach from ancestry side by state_transitioning, by requiring exact match on both root hash and slot for fast path
* make more precise the fast path condition
* redo logic to make uniform with BeaconChainDB; fix chronos deprecation warning
* revert not-working replacement of deprecated chronos futures `or`
* switch testnet1 to mainnet
* refactor --verify-finalization from run_node to only be invoked via eth2_network_simulation and local sim; clean up various shell issues
* remove hardcoded --verify-finalization in launch_local_testnet
- Jenkins: we don't need LOG_LEVEL=TRACE twice for the same binaries
- eth2_network_simulation: allow overriding NETWORK_TYPE and
BOOTSTRAP_NODE_NETWORK_TYPE from the environment
* re-enable test_interop based on zcli with 0.9.1 specs and update initialize_beacon_state_from_eth1(...) to 0.9.1
* switch many procs to funcs
* fix import os.nim instead; ospaths is deprecated [Deprecated] warnings
The number of user nodes is now specified with `USER_NODES`.
To make the instructions more stable, the "numeric id" of the user
nodes will be starting from 0 (so you can always use `run_node.sh 0`
to start a user node).
If you specify a node index above the total number of nodes, you'll
launch a node without any validators attached (this is useful for
testing the sync for example).
Multi-client testing requires more portable formats, and SSZ is
much better specified than our flavour of Json.
Tools like ncli and zcli can be now used to inspect the contents
of the SSZ files.
* stop double-counting attestation-topic attestations
* track attestations sent/received in grafana
* address suggestion to clarify sent attestation count being across all nodes
* add interop launcher scripts
* stick validator_keygen into beacon_node
* fix lmd ghost slot number on missing block
* use mocked eth1data when producing blocks
* use bls public key method for withdrawal credentials
* fix deposit domain
* prefer lowercase for a bunch of toHex
* build simulation binary in data folder to avoid data types confusion
* relax attestation validation when attestation is incoming but make it
more strict when adding to block
* share attestation validation logic between attestation pool and state
transition
* remove a bunch of redundant logging
* fix potential underflow in attestation delay checking
* fix committee used for attestation shard selection when attesting
* fix attestation data construction
* working justification, 1 node, 64 validators
* closer to tn1 params: 256 validators
* more debugging output; switch to minimum test case
* working justification and finalization in local network simulation
* fix currently incorrect state transition/attestation test assumption
The shell expansion syntax used in the run_node script was not handled properly
by the bash shell shipped with macOS. In particular, it was not able to handle
range expressions involving interpolated vars:
cp foo/{$FIRST_INDEX..$LAST_INDEX} ...
- updated README.md
- beacon_chain.nimble: accept compilation flags as params
- nim.cfg: enable `--opt:speed` at the top level
- simulation scripts: allow overriding GIT_ROOT and do some refactoring
* avoid crash on invalid block production (fixes#209)
* fetch blocks every second, roughly
* fix start.sh param order
* run beacon node sim at slower pace
* On the testnet site, the first N instead of the last N validators
will be considered reserved for users
* The number of validators is consistently named `totalValidators`
in all commands
* Proper keys are generated by default in validator_keygen
* add simple bitfield type (fixes#19)
* fix bit endianess to match spec
* consolidate attestation and block logging
* cruft cleanup
* run optimizer on tests, speeds up build
With these changes, running a simulation is very close to running
an actual testnet. Some checks have been added in the client to
make sure you are not connecting to an incompatible network (e.g.
a network running with a different number of shards).
Passing along no-name command line arguments when using make (from Nimbus) would need make target revamps that are a bit clumsy AFAIK, so using environment variables in the spirit of USE_MULTITAIL to get the number of validators and nodes makes more sense to me.
This lets us run the sim from nim-beacon-chain with
```bash
VALIDATORS=512 NODES=100 USE_MULTITAIL="yes" tests/simulation/start.sh
```
and from Nimbus with:
```bash
VALIDATORS=512 NODES=100 USE_MULTITAIL="yes" make eth2_network_simulation
```
Also makes for a more verbose launch process.
* set epoch_boundary_root - chain finalizes!
* fix slotStart to offset GENESIS_SLOT
* work around bug that will be fixed by
https://github.com/ethereum/eth2.0-specs/pull/732
* compile with debug info (there was a GC-related crash in C land)
* allow running more or fewer validators
* use deterministic key generation for tests to avoid exhausting system
RNG
* update README with simulator docs
* write the data of each validator to separate file, instead of a big
chainstart.json (makes it easier to run different validator counts)
* implement in-memory block graph
* store tail block in database
* resolve unknown parents by syncing them from peers
* introduce concept of resolved blocks and attestations - those that
follow minimal protocol rules
* update state head lazily
* log more stuff
* shortHash -> shortLog
* start 9/10 beacon nodes by default, last can be started manually
* see also #134
* fix start.sh epoch length
* fix initial attestation pool on reordered attestations
* simplify db layer api
* load head block from database on startup, then load state
* significantly changes database format
* move subscriptions to separate proc's
* implement block replay from historical state
* avoid rescheduling epoch actions on block receipt (why?)
* make sure genesis block is created and used
* relax initial state sim parameters a bit
* move attestation pool to separate file
* combine attestations lazily when needed
* advance state when there's a gap while attesting
* compile beacon node with optimizations - it's tooo slow right now
* log when unable to keep up
Ștefan Talpalaru
Zahary Karadjov
tersec
Dmitriy Ryajov
Jacek Sieka
Yuriy Glukhov
Yuriy Glukhov
Jacek Sieka
Bruno Škvorc
Mamy Ratsimbazafy