OpenShift provides some observability features out-of-the-box, but which features are available depends on what your cluster admin has working and what you have access to. On the NERC, we are able to see container metrics in OpenShift Developer's built-in dashboard. To enable custom visualizations, metrics analysis, and alerting, we need to connect this data to an instance of Grafana which we control.
Creating the superuser of a Django-based application is usually done by running the command
manage.py createsuperuser, hence it requires shell access. This makes sense as shell access
implies the person should also have admin privileges. However, shell access can be clunky
(think of how to run something in a container, kubectl get pods -n chris && kubectl exec -it -n chris <pod_name> python manage.py createsuperuser...).
We would prefer a declarative approach.
In our existing frontend application, React was utilized to create a Single Page Application (SPA), where pages were loaded and rendered exclusively in the browser. However, our more recent user interface (UI) has been developed using SvelteKit, which harnesses its Server-Side Rendering (SSR) capabilities.
Currently, https://cube.chrisproject.org is being powered by a VM called fnndsc.childrens.harvard.edu
in the Boston Children Hospital network's DMZ. It's been working well for us through
the years, however its 480GB disk frequently runs out of space. For more
storage, easier deployments, and stability, we want to try migrating this instance
of the ChRIS backend to the NERC's OpenShift cluster.
Currently, https://cube.chrisproject.org is being powered by a VM called fnndsc.childrens.harvard.edu
in the Boston Children Hospital network's DMZ. It's been working well for us through
the years, however its 480GB disk frequently runs out of space. For more
storage, we want to try migrating the data into NERC OpenStack Swift object storage.
$PATHLink to slides: intro_to_fnndsc_linux_aug2023.pdf
FNNDSC computer help and dev team email: Newborn_FNNDSCdev-dl@childrens.harvard.edu
As part of the Scientific Computing in Rust 2023 series, Jennings presents a lightning talk explaining his choice of the using the language Rust to develop a command-line client for the medical compute platform ChRIS. The language features of Rust, in addition to its rich ecosystem of high quality libraries, makes it a productive choice for developing CLI tools and prototyping experiments.
We were invited to present our efforts of getting ChRIS to run on Podman at the Podman Community meeting.
Raghuram Banda and Ishu Verma of Red Hat interview Jennings about how he is using ChRIS to power quantitative neuroanatomy research at the FNNDSC of the Boston Children's Hospital.