Facebook’s Calibra team explains a new method for BTF testing
- Calibra’s new method for BFT testing is d믭 Twins.
- Twins involves running two instances of a node with the same identity to emulate Byzantine behaviour.
- While Twins shows promise, it does not cover all Byzantine behaviours.
Facebook’s Calibra team has released a research paper that outlines a new technique for Byzantine Fault Tolerance (BFT) testing. Calibra’s innovation is dubbed “Twins” and it involves a lightweight criterion for testing a network’s BFT implementation. Per the publication, BFT systems have been studied extensively for more than 20 years. However, Calibra believes that the sector lacks a principled strategy for testing BFT implementations.
According to the research paper, the Twins approach involves running two instances of a node with the same identity to mirror Byzantine behavior. Calibra noted that Twins generates Byzantine behaviors such as equivocation, double voting, and losing internal state.
Are you looking for fast-news, hot-tips and market analysis? Sign-up for the Invezz newsletter, today.
This method also foregoes behaviors that honest nodes reject easily. An example of such behaviors is creating semantically invalid messages. Through this, Calibra believes that “Twins can systematically generate Byzantine attack scenarios at scale, execute them in a controlled manner, and check for desired protocol properties”.
The Twins method shows promise but has limitations
According to the Calibra team,
“Twins successfully reinstates several famous attacks on BFT protocols. In all cases, protocols break within fewer than a dozen protocol steps, hence it is realistic for the Twins approach to expose the problems. In two of these attacks, it took the community more than a decade to discover protocol flaws that Twins would have surfaced within minutes.”
Calibra added that Twins executed 3M Twins-generated scenarios, and exposed (self-injected) subtle safety bugs within minutes of testing after being introduced in a production setting.
However, Twins does not cover all Byzantine behaviors, according to the paper. These include behaviors that do not adhere to the full disclosure of causal past.
Commenting on this feat, Calibra co-creator, and board member, David Marcus tweeted,
“One of my greatest joys at work is getting to work with some of the brightest and most capable people in the industry. Here’s some cutting edge work on testing BFT implementations by our mighty research team.”
The BFT concept emanates from a 1982 paper authored by Leslie Lamport, Robert Shostak, and Marshall Pease. The paper describes a situation where Byzantine generals and their armies plan to lay siege to a castle. For a successful attack, all parties must work in tandem. However, the generals are aware of a traitor among them. As a result, it is hard to work as a unit.
Applied to the blockchain, BFT refers to the challenges that members of a network face while trying to work in unison amid malicious actors.
Do you think Twins will help minimize the number of malicious actors trying to cause harm to blockchain networks? Share your thoughts in the comment section below.