What the study found
The study reports that a fuzzy hybrid reptile–mamba optimisation framework for cloud orchestration performed better under correlated faults and bursty workloads. It treats resource allocation as a risk-constrained problem using a conditional value-at-risk (CVaR) penalty, which is a way to account for adverse outcomes in optimization.
Why the authors say this matters
The authors suggest the work matters because cloud platforms face reliability threats from correlated hardware faults, workload surges, and competing performance goals. The study suggests their framework can address these risks through fault-aware orchestration and self-healing control.
What the researchers tested
The researchers developed the fuzzy hybrid reptile–mamba optimisation (FHRMO) framework, combining a CVaR-penalised multi-objective model, a new black mamba local search operator, and an event-triggered self-healing policy. They tested it in CloudSim Plus using Google Cluster Trace data, and they evaluated it with Friedman and Wilcoxon tests over 30 independent runs across five stress scenarios.
What worked and what didn't
The abstract says the framework produced statistically significant gains across five stress scenarios, including correlated faults and bursty workloads. It also states that the self-healing policy had a local sufficient-decrease guarantee for repair cost and a hysteresis-based bounded-switching guarantee for the controller. The abstract does not identify any failed cases or specific metrics that did not improve.
What to keep in mind
The summary is based only on the abstract, so details of effect size, baseline comparisons, and scenario-specific numbers are not provided. Limitations are not described in the available abstract.
Key points
- The paper presents a fuzzy hybrid reptile–mamba optimisation framework for cloud orchestration.
- It uses a CVaR-penalised multi-objective model to balance fault tolerance, response latency, and recovery overhead.
- A new black mamba local search operator and an event-triggered self-healing policy are part of the framework.
- Tests in CloudSim Plus with Google Cluster Trace data showed statistically significant gains in five stress scenarios.
- The abstract does not report specific negative results or detailed limitations.
Disclosure
- Research title:
- Adaptive cloud orchestration improved resilience under correlated faults
- Authors:
- Michaelraj Kingston Roberts, Sampath Kumar Shanmugam, Sarah M. Alhammad, Doaa Sami Khafaga
- Institutions:
- Princess Nourah bint Abdulrahman University, Princess Nourah bint Abdulrahman University, Sri Eshwar College of Engineering, Sri Eshwar College of Engineering
- Publication date:
- 2026-04-02
- OpenAlex record:
- View
Get the weekly research newsletter
Stay current with scholarly research without reading academic papers — one filtered digest, every Friday.