go.dedis.ch/onet/v3
The Overlay-network (Onet) is a library for simulation and deployment of decentralized, distributed protocols. This library offers a framework for research, simulation, and deployment of crypto-related protocols with an emphasis on decentralized, distributed protocols. It offers an abstraction for tree-based communications between thousands of nodes and it is used both in research for testing out new protocols and running simulations, as well as in production to deploy those protocols as a service in a distributed manner.
Onet is developed by DEDIS/EFPL as part of the Cothority project that aims to deploy a large number of nodes for distributed signing and related projects. In cothority, nodes are commonly named "conodes". A collective authority (cothority) is a set of conodes that work together to handle a distributed, decentralized task.
Onet allows you to set up the following three elements:
protocols: a short-lived set of messages being passed back and forth between one or more conodes
services: define an API usable by client programs and instantiate protocols
apps: communicate with the service-API of one or more conodes
We also provide a set of methods to set up and run simulations.
:book: Table of Contents
app - useful libraries if you want to create a CLI app for the cothority
cfgpath - single package to get the configuration-path
log - everybody needs its own log-library - this one has log-levels, colors, time, ...
network - different type of connections: channels, tcp, tls
simul - allowing to run your protocols and services on different platforms with up to 50'000 nodes
The Onet library follows the same development cycle as the one described in the dedis/cothority project.
This project and all its files are licensed under the GNU Lesser General Public License, version 3 (SPDX short identifier: LGPL-3.0)
If this license is not suitable for your business or project please contact us for a full commercial license.
If you want to contribute, please have a look at CONTRIBUTION for licensing details and feel free to open a pull request.
You can contact us at https://groups.google.com/forum/#!forum/cothority or privately at dedis@epfl.ch.
This library is offered as-is without any guarantees. It would need an independent security review before it should be considered ready for use in security-critical applications. If you integrate Onet into your application it is YOUR RESPONSIBILITY to arrange for that audit.
If you notice a possible security problem, please report it to dedis-security@epfl.ch.
In Onet, you define Services that use Protocols which can send and receive messages. Each Protocol is instantiated when needed as a ProtocolInstance. As multiple Protocols can be run at the same time, there can be more than one ProtocolInstance of the same Protocol. Onet makes sure all messages get routed to the appropriate ProtocolInstance.
Foreign applications can communicate with Onet over the service-API, which is implemented using protobuf over WebSockets for JavaScript compatibility.
This chapter provides a high-level description of the cothority framework. Let's start with a picture and then dive into each main components of the library.
As you can see there's a bunch of different entities involved. Let's get down the rabbit hole to explain the most important ones!
The Router handles all incoming and outgoing messages from and to the network. A Router can use different underlying types of connection:
A conode is the main entity of a Cothority server. It holds the Router, the
Overlay, and the different Services. Generally, for developing an application
using the framework, you would create your Router first, then the Conode, and
then finally call conode.Start().
A Roster is simply a list of Conodes denoted by their public key and address. A Roster is identified by its ID, which is unique for each list.
A Protocol is an interface where users of the library must implement the logic of the protocol they want to code. It is supposed to be a short term entity that is self-sufficient, i.e. it does not need external access to any other resources of the Cothority framework. A protocol can be launched from SDA itself or by a Service.
A Service is a long term entity that is created when a Conode is created. It serves different purposes:
A ServiceManager is the main interface between the Conode and the Service. It transforms and embed the messages created by the Service to its own format and pass it to the Conode for the sending part; vice versa for the reception part.
A Tree is a standard tree data structure where each node - called TreeNode - is denoted by its public key and address. The Tree is constructed out of a Roster.
The Overlay provides an abstraction to communicate over different Trees that the Protocols and Services need. It handles the following tasks:
A TreeNodeInstance is created by the Overlay. There is one TreeNodeInstance for
each Protocol and it acts as the central point of communication for that
Protocol. The TreeNodeInstance offers to its Protocol some common tree methods
such as SendParent,SendChild, IsRoot etc. More importantly, it transforms
and embeds the message given by the Protocol into its own struct and dispatch it
to the Overlay for the sending part; vice versa for the reception part.
An application in the context of Onet is a CLI-program that interacts with one or more conodes through the use of the API defined by one or more services. It is mostly written in go, but in the cothority-repository you can also find libraries to interact in javascript and java.
Users of Onet have the option to make use of its built-in database.
We use bbolt, which supports "fully serializable ACID transactions" to ensure data integrity for Onet users. Users should be able to do the following:
Users are recommended to perform frequent backups such that data can be
recovered if Onet nodes fail. Onet stores all of its data in the context folder,
specified by $CONODE_SERVICE_PATH. If unset, it defaults to
~/Library/Application Support/conode/data on macOS,$HOME\AppData\Local\Conode on Windows, or~/.local/share/conode on other Unix/Linux.Hence, to backup, it is recommended to use a standard backup tool, such as rsync, and copy the folder to a different physical location periodically. The database keeps a transaction log.
Performing backups in the middle of a transaction should not be a problem.
However, it is still recommended to check the data integrity of the backed-up
file using the bbolt CLI, i.e. bolt check database_name.db.
To install the bbolt CLI, see Bolt Installation.
Data corruption is easy to detect as Onet nodes crash when reading from a
corrupted database, at startup or during operation. Concretely, the bbolt
library would panic
(source).
This behavior is produced by writing a few blocks of random data using dd to
the database.
In case of data corruption, the database must be restored from a backup by simply copying the backup copy to the context directory, and then starting the conode again. It is the user's responsibility to make sure that the data is up to date, e.g. by reading the latest data from running Onet nodes.
The primary and recommended methods to interact with the database are
Load and
Save. If more control
on the database is needed, then we can ask the context to return a database
handler and bucket name using the function
GetAdditionalBucket.
All the bbolt functions can be
used with the database handler. However, the user should avoid creating new
buckets using the bbolt functions and only use GetAdditionalBucket to avoid
bucket name conflicts.
Have a look at the simul/README.md for explanations about simulations.
FAQs
Unknown package
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Research
Security News
Socket’s threat research team has detected six malicious npm packages typosquatting popular libraries to insert SSH backdoors.
Security News
MITRE's 2024 CWE Top 25 highlights critical software vulnerabilities like XSS, SQL Injection, and CSRF, reflecting shifts due to a refined ranking methodology.
Security News
In this segment of the Risky Business podcast, Feross Aboukhadijeh and Patrick Gray discuss the challenges of tracking malware discovered in open source softare.