Nim implementation of the Ethereum Beacon Chain https://nimbus.guide
Go to file
Jacek Sieka 2081c4f505
avoid unnecessary seq allocations (#1573)
"sigh"
2020-08-27 08:32:51 +02:00
.github/ISSUE_TEMPLATE Update issue templates (#1488) 2020-08-11 18:21:12 +02:00
.vscode vscode: fix tasks 2019-11-25 13:51:47 +01:00
beacon_chain avoid unnecessary seq allocations (#1573) 2020-08-27 08:32:51 +02:00
benchmarks bump some eth2 phase0 spec refs to v0.11.1 2020-04-17 16:37:52 +03:00
docker entry_point.sh: account for local modifications 2020-08-04 03:12:42 +02:00
docs Medalla guide updates + troubleshooting page (#1571) 2020-08-26 19:27:07 +02:00
grafana update Grafana dashboard 2020-08-23 18:44:15 +02:00
media forgot the image 2020-06-11 16:00:56 +02:00
multinet Remove multinet scripts and forward to their new location [skip ci] 2019-10-29 19:18:01 +02:00
nbench use cache in process_voluntary_exit() (#1507) 2020-08-14 12:42:59 +00:00
ncli renamed CandidateChains to ChainDagRef and made the Quarantine type a ref type so there is a single instance in the beacon node (#1407) 2020-07-31 14:49:06 +00:00
nfuzz speed up shuffling 2020-08-21 16:05:10 +03:00
research rework epoch cache referencing 2020-08-19 10:09:06 +03:00
scripts run logtrace in finalization CI (#1512) 2020-08-17 08:36:29 +02:00
tests Fork choice - almost free pruning - fix #1534 (#1535) 2020-08-26 17:23:34 +02:00
tools run 6 nodes per testnet [skip ci] 2020-06-22 21:51:47 +02:00
vendor avoid unnecessary seq allocations (#1573) 2020-08-27 08:32:51 +02:00
wasm Mechanically remove all mentions of the daemon from the code 2020-03-23 14:59:33 +02:00
.appveyor.yml Travis: disable the ARM64 job (too unreliable) (#1276) 2020-07-03 10:04:04 +00:00
.editorconfig Makefile targets for ctail and ntu 2020-06-22 17:30:04 +03:00
.gitignore Move the nimbus books in a subfolder to allow for more books (#1263) 2020-07-01 10:37:04 +02:00
.gitmodules bump submodules (#1472) 2020-08-08 18:00:19 +02:00
.nvmrc Local sim now includes a prometheus server, a Ganache instance and a simulation dashboard server (in Tmux mode) 2020-06-22 17:30:04 +03:00
.travis.yml Bump nim-blscurve (#1491) 2020-08-15 19:33:58 +02:00
AllTests-mainnet.md rework epoch cache referencing 2020-08-19 10:09:06 +03:00
FixtureAll-mainnet.md add EF finality tests (#1515) 2020-08-17 07:19:48 +00:00
FixtureSSZConsensus-mainnet.md allow compile-time switching between v0.11.3 and v0.12.1 2020-06-09 18:39:05 +00:00
FixtureSSZGeneric-mainnet.md Reform the networking layer in order to handle the new stricter SSZ API 2020-05-24 19:00:34 +03:00
Jenkinsfile Jenkins: run benchmarks (#1524) 2020-08-26 17:31:46 +02:00
Jenkinsfile.benchmarks Jenkins: run benchmarks (#1524) 2020-08-26 17:31:46 +02:00
LICENSE-APACHEv2 initial commit 2018-07-20 15:46:03 +02:00
LICENSE-MIT initial commit 2018-07-20 15:46:03 +02:00
Makefile Add a CPU_LIMIT option for the Medalla and Altona targets [skip ci] 2020-08-22 01:50:38 +03:00
README.md Add a CPU_LIMIT option for the Medalla and Altona targets [skip ci] 2020-08-22 01:50:38 +03:00
azure-pipelines.yml increase Jenkins timeout from 90 to 100 minutes (#1519) 2020-08-18 07:13:53 +00:00
beacon_chain.nimble Bump nim-blscurve (#1491) 2020-08-15 19:33:58 +02:00
config.nims Bump nim-blscurve (#1491) 2020-08-15 19:33:58 +02:00
env.sh Local sim now includes a prometheus server, a Ganache instance and a simulation dashboard server (in Tmux mode) 2020-06-22 17:30:04 +03:00

README.md

Nimbus Eth2 (Beacon Chain)

Build Status (Travis) Build Status (Azure) License: Apache License: MIT Stability: experimental

Discord: Nimbus Status: #nimbus-general

Nimbus beacon chain is a research implementation of the beacon chain component of the upcoming Ethereum Serenity upgrade, aka Eth2.

Documentation

You can find the information you need to run a beacon node and operate as a validator in The Book.

See this page for how to become a validator on Medalla.

You can check where the beacon chain fits in the Ethereum ecosystem our Two-Point-Oh series: https://our.status.im/tag/two-point-oh/

Prerequisites for everyone

At the moment, Nimbus has to be built from source.

Nimbus has the following external dependencies:

  • Developer tools (C compiler, Make, Bash, Git)
  • PCRE

Nim is not an external dependency, Nimbus will build its own local copy.

Linux

On common Linux distributions the dependencies can be installed with:

# Debian and Ubuntu
sudo apt-get install build-essential git libpcre3-dev

# Fedora
dnf install @development-tools pcre

# Archlinux, using an AUR manager for pcre-static
yourAURmanager -S base-devel pcre-static

MacOS

Assuming you use Homebrew to manage packages:

brew install pcre cmake

Make sure you have CMake installed, to be able to build libunwind (used for lightweight stack traces).

Windows

You can install the developer tools by following the instruction in our Windows dev environment section. It also provides a downloading script for prebuilt PCRE.

Android

  • Install the Termux app from FDroid or the Google Play store
  • Install a PRoot of your choice following the instructions for your preferred distribution. Note, the Ubuntu PRoot is known to contain all Nimbus prerequisites compiled on Arm64 architecture (common architecture for Android devices).

Assuming Ubuntu PRoot is used

apt install build-essential git libpcre3-dev

For users

Connecting to testnets

Once the prerequisites are installed you can connect to the Medalla testnet with the following commands:

git clone https://github.com/status-im/nim-beacon-chain
cd nim-beacon-chain
make medalla           # This will build Nimbus and all other dependencies
                      # and connect you to Medalla

You can also start multiple local nodes, in different terminal windows/tabs, by specifying their numeric IDs:

make medalla NODE_ID=0 # the default
make medalla NODE_ID=1
make medalla NODE_ID=2

To change the TCP and UDP ports from the default value of 9000:

make BASE_PORT=9100 medalla

If you wish to make a deposit, execute the following command:

make medalla-deposit VALIDATORS=2 # The default is just 1 deposit

Getting metrics from a local testnet client

# the primitive HTTP server started to serve the metrics is considered insecure
make NIMFLAGS="-d:insecure" medalla

You can now see the raw metrics on http://127.0.0.1:8008/metrics but they're not very useful like this, so let's feed them to a Prometheus instance:

cd build/data/shared_medalla_0
prometheus --config.file=./prometheus.yml --storage.tsdb.path=./prometheus
# when starting multiple nodes at the same time, just use the config file from the one with the highest ID

For some pretty pictures, get Grafana up and running, then import the dashboard definition in "grafana/beacon_nodes_Grafana_dashboard.json".

Stress-testing the client by limiting the CPU power

make medalla CPU_LIMIT=20

The limiting is provided by the cpulimit utility, available on Linux and macOS. The specified value is a percentage of a single CPU core. Usually 1 - 100, but can be higher on multi-core CPUs.

Interop (for other Eth2 clients)

After installing the prerequisites

To run the Nimbus state transition, we provide the ncli tool:

The interop scripts have been moved in a common repo, the interop relied on 0.8.3 specs which had seen significant changes. The interop branch still exist but is unmaintained.

For researchers

State transition simulation

The state transition simulator can quickly run the Beacon chain state transition function in isolation and output JSON snapshots of the state. The simulation runs without networking and blocks are processed without slot time delays.

# build and run the state simulator, then display its help ("-d:release" speeds it
# up substantially, allowing the simulation of longer runs in reasonable time)
make NIMFLAGS="-d:release" state_sim
build/state_sim --help

Local network simulation

The local network simulation will create a full peer-to-peer network of beacon nodes and validators on a single machine, and run the beacon chain in real time.

Parameters such as shard, validator counts, and data folders are configured vars.sh. They can be set in as environment variables before launching the simulation.

# Clear data files from your last run and start the simulation with a new genesis block:
make VALIDATORS=192 NODES=6 USER_NODES=1 eth2_network_simulation

# In another terminal, get a shell with the right environment variables set:
./env.sh bash

# In the above example, the network is prepared for 7 beacon nodes but one of
# them is not started by default (`USER_NODES`) - this is useful to test
# catching up to the consensus. The following command will start the missing node.
./tests/simulation/run_node.sh 0 # (or the index (0-based) of the missing node)

# Running a separate node allows you to test sync as well as see what the action
# looks like from a single nodes' perspective.

By default, all validators are loaded within the beacon nodes, but if you want to use external processes as validator clients you can pass BN_VC_VALIDATOR_SPLIT=yes as an additional argument to the make eth2_network_simulation command and that will split the VALIDATORS between beacon nodes and validator clients - for example with 192 validators and 6 nodes you will end up with 6 beacon node and 6 validator client processes, where each of them will handle 16 validators.

By default, the simulation will start from a pre-generated genesis state. If you wish to simulate the bootstrap process with a Ethereum 1.0 validator deposit contract, start the simulation with WAIT_GENESIS=yes

make eth2_network_simulation WAIT_GENESIS=yes

You can also separate the output from each beacon node in its own panel, using multitail:

make eth2_network_simulation USE_MULTITAIL="yes"

You can find out more about it in the development update.

Alternatively, fire up our experimental Vagrant instance with Nim pre-installed and give us yout feedback about the process!

Visualising simulation metrics

The generic instructions from the Nimbus repo apply here as well.

Specific steps:

# This will generate the Prometheus config on the fly, based on the number of
# nodes (which you can control by passing something like NODES=6 to `make`).
# The `-d:insecure` flag starts an HTTP server from which the Prometheus daemon will pull the metrics.
make VALIDATORS=192 NODES=6 USER_NODES=0 NIMFLAGS="-d:insecure" eth2_network_simulation

# In another terminal tab, after the sim started:
cd tests/simulation/prometheus
prometheus

The dashboard you need to import in Grafana is "grafana/beacon_nodes_Grafana_dashboard.json".

monitoring dashboard

Network inspection

The inspector tool can help monitor the libp2p network and the various channels where blocks and attestations are being transmitted, showing message and connectivity metadata. By default, it will monitor all ethereum 2 gossip traffic.

. ./env.sh
# Build inspector for minimal config:
./env.sh nim c -d:const_preset=minimal -o:build/inspector_minimal beacon_chain/inspector.nim

# Build inspector for mainnet config:
./env.sh nim c -d:const_preset=mainnet -o:build/inspector_mainnet beacon_chain/inspector.nim

# See available options
./env.sh build/inspector_minimal --help

# Connect to a network from eth2 testnet repo bootstrap file - --decode option attempts to decode the messages as well
./env.sh build/inspector_minimal --decode -b:$(curl -s https://raw.githubusercontent.com/eth2-clients/eth2-testnets/master/nimbus/testnet0/bootstrap_nodes.txt | head -n1)

For developers

Latest updates happen in the devel branch which is merged into master every week on Tuesday before deploying new testnets.

Interesting Make variables and targets are documented in the nimbus-build-system repo.

The following sections explain how to set up your build environment on your platform.

Windows dev environment

Install Mingw-w64 for your architecture using the "MinGW-W64 Online Installer" (first link under the directory listing). Run it and select your architecture in the setup menu ("i686" on 32-bit, "x86_64" on 64-bit), set the threads to "win32" and the exceptions to "dwarf" on 32-bit and "seh" on 64-bit. Change the installation directory to "C:\mingw-w64" and add it to your system PATH in "My Computer"/"This PC" -> Properties -> Advanced system settings -> Environment Variables -> Path -> Edit -> New -> C:\mingw-w64\mingw64\bin (it's "C:\mingw-w64\mingw32\bin" on 32-bit)

Install Git for Windows and use a "Git Bash" shell to clone and build nim-beacon-chain.

Install CMake to be able to build libunwind (used for lightweight stack traces).

When running the tests, you might hit some Windows path length limits. Increase them by editing the Registry in a PowerShell instance with administrator privileges:

Set-ItemProperty -Path 'HKLM:\SYSTEM\CurrentControlSet\Control\FileSystem' -Name 'LongPathsEnabled' -Value 1

and run this in a "Git Bash" terminal:

git config --global core.longpaths true

If you don't want to compile PCRE separately, you can fetch pre-compiled DLLs with:

mingw32-make fetch-dlls # this will place the right DLLs for your architecture in the "build/" directory

If you were following the Windows testnet instructions, you can jump back to Connecting to testnets now

You can now follow those instructions in the previous section by replacing make with mingw32-make (regardless of your 32-bit or 64-bit architecture):

mingw32-make test # run the test suite

Linux, MacOS

After cloning the repo:

# The first `make` invocation will update all Git submodules.
# You'll run `make update` after each `git pull`, in the future, to keep those submodules up to date.

# Build beacon_node and all the tools, using 4 parallel Make jobs
make -j4

# Run tests
make test

# Update to latest version
git pull
make update

To run a command that might use binaries from the Status Nim fork:

./env.sh bash # start a new interactive shell with the right env vars set
which nim
nim --version # Nimbus is tested and supported on 1.0.2 at the moment

# or without starting a new interactive shell:
./env.sh which nim
./env.sh nim --version

Raspberry Pi

We recommend you remove any cover or use a fan; the Raspberry Pi will get hot (85°C) and throttle.

  • Raspberry PI 3b+ or Raspberry Pi 4b.
  • 64gb SD Card (less might work too, but the default recommended 4-8GB will probably be too small)
  • Rasbian Buster Lite - Lite version is enough to get going and will save some disk space!

Assuming you're working with a freshly written image:


# Start by increasing swap size to 2gb:
sudo vi /etc/dphys-swapfile
# Set CONF_SWAPSIZE=2048
# :wq
sudo reboot

# Install prerequisites
sudo apt-get install git libgflags-dev libsnappy-dev libpcre3-dev

# Then you can follow instructions for Linux.

Makefile tips and tricks for developers

  • build all those tools known to the Makefile:
# $(nproc) corresponds to the number of cores you have
make -j$(nproc)
  • build a specific tool:
make state_sim
  • you can control the Makefile's verbosity with the V variable (defaults to 0):
make V=1 # verbose
make V=2 test # even more verbose
make LOG_LEVEL=DEBUG bench_bls_sig_agggregation # this is the default
make LOG_LEVEL=TRACE beacon_node # log everything
  • pass arbitrary parameters to the Nim compiler:
make NIMFLAGS="-d:release"
  • you can freely combine those variables on the make command line:
make -j$(nproc) NIMFLAGS="-d:release" USE_MULTITAIL=yes eth2_network_simulation
make USE_LIBBACKTRACE=0 # expect the resulting binaries to be 2-3 times slower
  • publish a book using mdBook from sources in "docs/" to GitHub pages:
make publish-book

CI setup

Local testnets run for 4 epochs each, to test finalization. That happens only on Jenkins Linux hosts, and their logs are available for download as artifacts, from the job's page. Don't expect these artifacts to be kept more than a day after the corresponding branch is deleted.

Jenkins artifacts

License

Licensed and distributed under either of

or

at your option. These files may not be copied, modified, or distributed except according to those terms.