ca61a7009d
* Bump submodules and related fixes * Fix some warnings due to chronos and Nim bumps * Seems macOS and Windows have different behavior |
||
---|---|---|
.. | ||
common | ||
docs | ||
eth_data | ||
grafana | ||
network | ||
network_data/testnet0 | ||
rpc | ||
scripts | ||
tests | ||
tools | ||
README.md | ||
conf.nim | ||
content_db.nim | ||
fluffy.nim | ||
fluffy.nim.cfg | ||
logging.nim | ||
network_metadata.nim | ||
nim.cfg | ||
seed_db.nim | ||
version.nim |
README.md
Fluffy: The Nimbus Portal Network Client
Introduction
This folder holds the development of the Nimbus client implementation supporting the Portal Network: Fluffy. The Portal Network is a project still heavily in research phase and fully in flux. This client is thus still highly experimental.
Current status of specifications can be found in the portal-network-specs repository.
Development Updates
Monthly development updates are shared here.
To keep up to date with changes and development progress, follow the Nimbus blog.
How to Build & Run
Prerequisites
- GNU Make, Bash and the usual POSIX utilities. Git 2.9.4 or newer.
Build the Fluffy client
git clone git@github.com:status-im/nimbus-eth1.git
cd nimbus-eth1
make fluffy
# See available command line options
./build/fluffy --help
# Example command: Run the client and connect to a bootstrap node.
./build/fluffy --bootstrap-node:enr:<base64 encoding of ENR>
Update and rebuild the Fluffy client
# From the nimbus-eth1 repository
git pull
# To bring the git submodules up to date
make update
make fluffy
Run a Fluffy client on the public testnet
# Connect to the Portal testnet bootstrap nodes and enable the JSON-RPC APIs
./build/fluffy --rpc --table-ip-limit:1024 --bucket-ip-limit:24
The table-ip-limit
and bucket-ip-limit
options are needed to allow more
nodes with the same IPs in the routing tables. The default limits are there
as security measure. It is currently needed to increase the limits for the testnet
because the fleet of Fluffy nodes runs on only 2 machines / network interfaces.
There is a public Portal testnet which contains nodes of different clients.
Fluffy will default join the network through these bootstrap nodes.
You can also explicitly provide the --network:testnet0
option to join this
network, or --network:none
to not connect to any of these bootstrapn odes.
Note: The
--network
option selects automatically a static set of specific bootstrap nodes belonging to a "testnet". Currentlytestnet0
is the only option, which results in connecting to the testnet bootstrap nodes. It should be noted that there is currently no real way to distinguish a "specific" Portal network, and as long as the same Portal protocols are supported, nodes can simply connect to it and no real separation can be made. When testing locally the--network:none
option can be provided to avoid connecting to any of the testnet bootstrap nodes.
The Portal testnet is slowly being filled up with historical data through bridge nodes.
Because of this, more recent history data is more likely to be available.. This can
be tested by using the JSON-RPC call eth_getBlockByHash
:
# Get the hash of a block from your favorite block explorer, e.g.:
BLOCKHASH=0x34eea44911b19f9aa8c72f69bdcbda3ed933b11a940511b6f3f58a87427231fb # Replace this to the block hash of your choice
# Run this command to get this block:
curl -s -X POST -H 'Content-Type: application/json' -d '{"jsonrpc":"2.0","id":"1","method":"eth_getBlockByHash","params":["'${BLOCKHASH}'", true]}' http://localhost:8545 | jq
One can also use the blockwalk
tool to walk down the blocks one by one, e.g:
make blockwalk
BLOCKHASH=0x34eea44911b19f9aa8c72f69bdcbda3ed933b11a940511b6f3f58a87427231fb # Replace this to the block hash of your choice
./build/blockwalk --block-hash:${BLOCKHASH}
Run Fluffy test suite
# From the nimbus-eth1 repository
make fluffy-test
Run Fluffy local testnet script
./fluffy/scripts/launch_local_testnet.sh
Find more details on the usage and workings of the local testnet script here.
Windows support
Follow the steps outlined here to build fluffy on Windows.
Development tools and documentation
The fluffy directory also holds several tools to help development of the Portal networks.
Command to build the tools:
make fluffy-tools -j6
Additional documention on the tools or on what you can use them for:
- eth_data_exporter: tool to extract content from EL or CL and prepare it as Portal content and content keys.
- Content seeding: Documentation on how to retrieve & generate history data and how to seed it into the network
- Manual protocol interop testing: commands on how to manually test the discv5 and Portal protocol request and responses
- Local testnet script: Documentation on the local testnet script and how to use it
The basics for developers
When working on this repository, you can run the env.sh
script to run a
command with the right environment variables set. This means the vendored
Nim and Nim modules will be used, just as when you use make
.
E.g.:
# start a new interactive shell with the right env vars set
./env.sh bash
More development tips can be found on the general nimbus-eth1 readme.
The code follows the Status Nim Style Guide.
Build local dev container for portal-hive
To develop code against portal-hive tests you will need:
-
Clone and build portal-hive (#1)
-
Modify
Dockerfile
for fluffy inportal-hive/clients/fluffy/Dockerfile
(#2) -
Build local dev container using following command:
docker build --tag fluffy-dev --file ./fluffy/tools/docker/Dockerfile.portalhive .
You may need to change fluffy-dev to the tag you using in portal-hive client dockerfile. -
Run the tests
Also keep in mind that ./vendors
is dockerignored and cached. If you have to make local changes to one of the dependencies in that directory you'll have to remove vendors/
from ./fluffy/tools/docker/Dockerfile.portalhive.dockerignore
.
Metrics and their visualisation
To enable metrics run Fluffy with the --metrics
flag:
./build/fluffy --metrics
Default the metrics are available at http://127.0.0.1:8008/metrics.
The address can be changed with the --metrics-address
and --metrics-port
options.
This provides only a snapshot of the current metrics. In order track the metrics over time and to also visualise them one can use for example Prometheus and Grafana.
The steps on how to set up such system is explained in this guide.
A Fluffy specific dashboard can be found here.
This is the dashboard used for our Fluffy testnet fleet.
In order to use it locally, you will have to remove the
{job="nimbus-fluffy-metrics"}
part from the instance
and container
variables queries in the dashboard settings. Or they can also be changed to a
constant value.
The other option would be to remove those variables and remove their usage in each panel query.
License
Licensed and distributed under either of
- MIT license: LICENSE-MIT or http://opensource.org/licenses/MIT
or
- Apache License, Version 2.0, (LICENSE-APACHEv2 or http://www.apache.org/licenses/LICENSE-2.0)
at your option. These files may not be copied, modified, or distributed except according to those terms.