Han L1, De La Pierre M1, Golovina E1, Blasco J1
1Do It Now, , Aotearoa / New Zealand
Biography:
Junior Computational Scientist at Do IT Now (Auckland, New Zealand). She works on research computing infrastructure and HPC system integration, including software deployment with EasyBuild and EESSI, Open OnDemand deployment and application development, and operational support across HPC environments. Her work focuses on automation and infrastructure-as-code practices to enhance usability and efficiency across HPC platforms.
Abstract:
Open OnDemand (OOD) has become a key access point for the HPC and research community, providing a single web-based interface for computational resources, applications, files, jobs, and interactive services. In this talk, we will explore how OOD can evolve into a better-connected platform for research computing. We will focus on two practical OOD integrations that improve connectivity, consistency, and user experience across HPC environments.
The first integration connects OOD with the European Environment for Scientific Software Installations (EESSI), a shared and optimised software stack designed to run consistently across different platforms. This integration allows users to select software modules and launch a remote desktop session. When the session starts, the selected software becomes available and ready to use. This reduces manual setup and supports reproducible workflows for training, testing, benchmarking, and software sharing across research centres.
The second integration connects OOD to a shared Grafana instance in a Grafana Viewer application, allowing users to access monitoring dashboards directly from the OOD without separate logins or exposing Grafana externally. Access is controlled through Grafana RBAC and LDAP mappings, ensuring users only see dashboards relevant to their role, project, or team. This provides a secure way to expose system metrics, job monitoring, service status, and operational views to users, support teams, and administrators.
Together, these integrations show how OOD can evolve into a flexible research computing hub that brings together scientific software delivery, monitoring, and secure service access within a single user-facing portal.
