ROSETTA-VERGE IS CONSIDERED ALPHA SOFTWARE. USE AT YOUR OWN RISK! COINBASE ASSUMES NO RESPONSIBILITY NOR LIABILITY IF THERE IS A BUG IN THIS IMPLEMENTATION.
rosetta-verge
provides a reference implementation of the Rosetta API for
Verge in Golang. If you haven't heard of the Rosetta API, you can find more
information here.
- Rosetta API implementation (both Data API and Construction API)
- UTXO cache for all accounts (accessible using
/account/balance
) - Stateless, offline, curve-based transaction construction from any SegWit-Bech32 Address
As specified in the Rosetta API Principles,
all Rosetta implementations must be deployable via Docker and support running via either an
online
or offline
mode.
YOU MUST INSTALL DOCKER FOR THE FOLLOWING INSTRUCTIONS TO WORK. YOU CAN DOWNLOAD DOCKER HERE.
Running the following commands will create a Docker image called rosetta-verge:latest
.
To download the pre-built Docker image from the latest release, run:
curl -sSfL https://raw.githubusercontent.com/vergecurrency/rosetta-verge/master/install.sh | sh -s
Do not try to install rosetta-verge using GitHub Packages!
After cloning this repository, run:
make build-local
Running the following commands will start a Docker container in
detached mode with
a data directory at <working directory>/verge-data
and the Rosetta API accessible
at port 8080
.
docker run -d --rm --ulimit "nofile=100000:100000" -v "$(pwd)/verge-data:/data" -e "MODE=ONLINE" -e "NETWORK=MAINNET" -e "PORT=8080" -p 8080:8080 -p 21102:21102 rosetta-verge:latest
If you cloned the repository, you can run make run-mainnet-online
.
docker run -d --rm -e "MODE=OFFLINE" -e "NETWORK=MAINNET" -e "PORT=8081" -p 8081:8081 rosetta-verge:latest
If you cloned the repository, you can run make run-mainnet-offline
.
docker run -d --rm --ulimit "nofile=100000:100000" -v "$(pwd)/verge-data:/data" -e "MODE=ONLINE" -e "NETWORK=TESTNET" -e "PORT=8080" -p 8080:8080 -p 21104:21104 rosetta-verge:latest
If you cloned the repository, you can run make run-testnet-online
.
docker run -d --rm -e "MODE=OFFLINE" -e "NETWORK=TESTNET" -e "PORT=8081" -p 8081:8081 rosetta-verge:latest
If you cloned the repository, you can run make run-testnet-offline
.
rosetta-verge
has been tested on an AWS c5.2xlarge instance.
This instance type has 8 vCPU and 16 GB of RAM.
To increase the load rosetta-verge
can handle, it is recommended to tune your OS
settings to allow for more connections. On a linux-based OS, you can run the following
commands (source):
sysctl -w net.ipv4.tcp_tw_reuse=1
sysctl -w net.core.rmem_max=16777216
sysctl -w net.core.wmem_max=16777216
sysctl -w net.ipv4.tcp_max_syn_backlog=10000
sysctl -w net.core.somaxconn=10000
sysctl -p (when done)
We have not tested rosetta-verge
with net.ipv4.tcp_tw_recycle
and do not recommend
enabling it.
You should also modify your open file settings to 100000
. This can be done on a linux-based OS
with the command: ulimit -n 100000
.
rosetta-verge
uses memory-mapped files to
persist data in the indexer
. As a result, you must run rosetta-verge
on a 64-bit
architecture (the virtual address space easily exceeds 100s of GBs).
If you receive a kernel OOM, you may need to increase the allocated size of swap space on your OS. There is a great tutorial for how to do this on Linux here.
rosetta-verge
uses the syncer
, storage
, parser
, and server
package
from rosetta-sdk-go
instead
of a new Verge-specific implementation of packages of similar functionality. Below
you can find a high-level overview of how everything fits together:
+------------------------------------------------------------------+
| |
| +--------------------------------------+ |
| | | |
| | indexer | |
| | | |
| | +--------+ | |
+-------------------+ pruner <----------+ | |
| | +--------+ | | |
+-----v----+ | | | |
| verged | | +------+--------+ | |
+-----+----+ | +--------> block_storage <----+ | |
| | | +---------------+ | | |
| | +---+----+ | | |
+-------------------> syncer | | | |
| +---+----+ | | |
| | +--------------+ | | |
| +--------> coin_storage | | | |
| +------^-------+ | | |
| | | | |
+--------------------------------------+ |
| | |
+-------------------------------------------------------------------------------------------+ |
| | | | |
| +------------------------------------------------------------+ | | |
| | | | |
| | +---------------------+-----------------------+------+ | |
| | | | | | |
| +-------+---------+ +-------+---------+ +-------+-------+ +-----------+----------+ | |
| | account_service | | network_service | | block_service | | construction_service +--------+
| +-----------------+ +-----------------+ +---------------+ +----------------------+ |
| |
| server |
| |
+-------------------------------------------------------------------------------------------+
- Automatically prune verged while indexing blocks
- Reduce sync time with concurrent block indexing
- Use Zstandard compression to reduce the size of data stored on disk without needing to write a manual byte-level encoding
To speed up indexing, rosetta-verge
uses concurrent block processing
with a "wait free" design (using channels instead of sleeps to signal
which threads are unblocked). This allows rosetta-verge
to fetch
multiple inputs from disk while it waits for inputs that appeared
in recently processed blocks to save to disk.
+----------+
| verged |
+-----+----+
|
|
+---------+ fetch block data / unpopulated txs |
| block 1 <------------------------------------+
+---------+ |
+--> tx 1 | |
| +---------+ |
| | tx 2 | |
| +----+----+ |
| | |
| | +---------+ |
| | | block 2 <-------------------+
| | +---------+ |
| +-----------> tx 3 +--+ |
| +---------+ | |
+-------------------> tx 4 | | |
| +---------+ | |
| | |
| retrieve previously synced | +---------+ |
| inputs needed for future | | block 3 <--+
| blocks while waiting for | +---------+
| populated blocks to save to +---> tx 5 |
| disk +---------+
+------------------------------------> tx 6 |
| +---------+
|
|
+------+--------+
| coin_storage |
+---------------+
To validate rosetta-verge
, install rosetta-cli
and run one of the following commands:
rosetta-cli check:data --configuration-file rosetta-cli-conf/testnet/config.json
rosetta-cli check:construction --configuration-file rosetta-cli-conf/testnet/config.json
rosetta-cli check:data --configuration-file rosetta-cli-conf/mainnet/config.json
- Publish benchamrks for sync speed, storage usage, and load testing
- Rosetta API
/mempool/transaction
implementation - Add CI test using
rosetta-cli
to run on each PR (likely on a regtest network) - Add performance mode to use unlimited RAM (implementation currently optimized to use <= 16 GB of RAM)
- Support Multi-Sig Sends
Please reach out on our community if you want to tackle anything on this list!
make deps
to install dependenciesmake test
to run testsmake lint
to lint the source codemake salus
to check for security concernsmake build-local
to build a Docker image from the local contextmake coverage-local
to generate a coverage report
This project is available open source under the terms of the Apache 2.0 License.
© 2021 Coinbase