Research
Here you will find references for academic projects, papers, and the research interests of Pystol.
Articles
- Chaos as a Software Product Line
A platform for improving open hybrid-cloud systems resiliency - Chaos in the core
Improving software-defined and cloud-native systems resiliency for 5G core networks
Chaos as a Software Product Line
A platform for improving open hybrid-cloud systems resiliency
Status: Submitted to Transactions on cloud computing ‘Under review’
Chaos in the core
Improving software-defined and cloud-native systems resiliency for 5G core networks
Status: Under development
Research interests
- Study how fault injection actions affect the system behavior deployed over large-scale cluster configurations.
- Stablish relationships between the injected faults, the size of the system and the impact on the system performance.
- Study the behavior of cloud-native systems with low latency requirements (for instance IoT and TelCo use cases), service mesh and software-defined networking applications, and edge architectures.
- Study backup, restore, and disaster recovery mechanisms after executing the fault injection actions. In this way, it is possible to recover the cluster consistent state in a shorter period of time and guarantee a reduction of the fault injection actions impact.
- Investigate different real-world applications to generate different workloads to be processed by the system, like cpu-intensive or data-intensive applications, produce different results when the fault actions are executed.
- Improve the observability of the system when fault injection actions are executed.