The objective of this thesis is to design and implement the deployment of a scalable search ecosystem for molecular dynamics (MD) datasets in a Kubernetes environment. While sharing research data alongside publications is becoming increasingly common, the field of molecular dynamics is beginning to follow this trend as well. MD simulations generate vast amounts of valuable data and results, and one of the most accessible ways for researchers to interact with such massive collections is through a simple search interface.

Behind such an interface, however, lies a complex data infrastructure composed of multiple components, capable of searching billions of MD trajectory snapshots within structural datasets in only a few seconds.

The focus of this thesis will be on deploying the scaled components of such an architecture in Kubernetes, ensuring efficient operation and resource usage as well as providing developers with streamlined development process.

All individual components, as well as the entire search ecosystem, should be deployable and configurable as Helm charts, providing a reproducible and maintainable deployment pipeline for large-scale MD search infrastructure.