Cross-organizational distributed ledgers (blockchains), such as Hyperledger Fabric, generally provide high integrity - but their dependability from other aspects (e.g, availability, reliability) is rather sitational and needs proper design as well as validation. The goal of the topic is to create a fault injection testbed for Hyperledger Fabric that enables subjecting a test system to a barrage of synthetically injected faults and analysing the effects. The secondary use case is chaos engineering - subjecting production, or at least representative CI/CD system instances to low-frequency faults to continuously validate fault tolerance and resilience on an ongoing basis.

This is a blockchain as well as DevOps topic, meaning that it's technically rather complex and requires a student who is already familiar with at least the basic tools of the trade (containerization, CI/CD solutions, preferably at least some Kubernetes) - Hyperledger Fabric in itself will require significant dedicated study time.