How can we improve blockchain systems to deal with unpredictable failures?

Trias
3 min readJun 8, 2022

--

TL; DR

In order to adapt to the development of blockchain in the crypto world, the design of blockchain system architecture is becoming increasingly complex, and the resulting problem is the unpredictability and difficulty in locating, which appear in case of system failures. Triathon Chaos Test Platform provides new-type and efficient test prevention that enables blockchain systems to effectively deal with unpredictable failures and improve the security and reliability of the system.

In the context of the proliferation of users and traffic, the market has put forward higher requirements for the performance of blockchain systems, and a growing number of project parties have started to try various methods to improve system performance, and further adapt to the current trends of the crypto world.

However, as the system architecture design becomes increasingly complex, the stability and reliability of the system are also being tested. For example, the interdependence of various components in the system and the crisscrossing of data access perplex the system’s operation links, which makes it difficult to find and locate failures.

Therefore, traditional testing, such as functional testing, performance testing, and high concurrency testing, these conventional testing methods gradually goes beyond finding defects in complex systems but requires a new testing method to find and fix the problems existing in complex systems.

Triathon Chaos Test Platform is a testing platform based on the theory “Chaos Engineering“, providing abundant testing cases, which offers system testing for blockchain platforms, supporting system failure simulation and analysis of multi-scenario and full-link, which would help projects improve system stability and reliability.

In a gigantic blockchain system, the emergence and impact of individual failure become unpredictable due to the complex connections among components, so the mindset of coping with failures should change from after-the-fact resolution to ex-ante prevention, i.e., identifying as many risks as possible before these failures or anomalies are triggered, and then conducting reinforcement and prevention purposefully so as to avoid serious consequences when failures occur.

Chaos testing is precisely such a methodology that actively identifies vulnerabilities in a system through continuous testing. Chaos testing is able to identify possible problems and risks in systems or during application by injecting various failure testing cases into the existing operating system within a controlled range.

Triathon Chaos Testing Platform is a community-driven testing shooting range that allows users to use different testing cases (Ship NFTs) in unordered combinations, to simulate the methods of chaos testing for public chains or others, so as to find problems in the system early and further help project teams optimize the system and ensure its safety and stability.

Users can participate in the process of chaos testing by pledging Ship NFTs. Different Ship NFTs represent different testing cases, such as stress tests, spike impact tests, vulnerability attack tests, etc. After users finish staking, the system will also randomly add other major obstacle testing cases in the test and disrupt the order of testing cases to simulate the actual scenario of failure occurrence to obtain the most realistic test data. After the test is completed, based on the performance of the staked testing cases in the testing process, users will be rewarded with GEON token as incentives including basic rewards, lucky rewards, and recording rewards and so on.

In addition, Triathon also allows users to deploy their own testing code to generate some special Ship NFTs, which can also be staked to the testing process for GEON rewards, through such a mechanism more users would be motivated to provide more test cases for the platform, eventually enriching the platform’s code base and enhancing the depth and breadth of testing.

--

--

Trias

Trustworthy and Reliable Intelligent Autonomous Systems