Kubernetes

TL;DR¶

To install DBRepo in your existing cluster, download the sample values.yaml for your deployment and update the variables, especially hostname.

helm upgrade --install dbrepo \
  -n dbrepo \
  "oci://registry.datalab.tuwien.ac.at/dbrepo/helm/dbrepo" \
  --values ./values.yaml \
  --version "1.7.3" \
  --create-namespace \
  --cleanup-on-fail

Prerequisites¶

Kubernetes 1.30+
PV provisioner support in the underlying infrastructure

Resource Quota¶

By default, any service requests a minimum quota of cpu and memory by default. These values are based on the absolute minimum needed to start the service. From our experience your cluster needs to meet the following ResourceQuota:

requests.cpu: 12000m
requests.ephemeral-storage: 3072Mi
secrets: '50'
<storageClass>.storageclass.storage.k8s.io/persistentvolumeclaims: '20'
<storageClass>.storageclass.storage.k8s.io/requests.storage: 150Gi
persistentvolumeclaims: '20'
requests.memory: 14336Mi
pods: '40'
requests.storage: 150Gi
limits.memory: 20480Mi
configmaps: '20'
services: '40'

You can reduce the resources needed by disabling services and metric collection. You can disable e.g. the dashboard service:

values.yaml

dashboardservice:
  enabled: false

Limitations¶

The entire Helm deployment is rootless (=runAsNonRoot=true) except for the Storage Service which still requires a root user.

Do you miss functionality? Do these limitations affect you?

We strongly encourage you to help us implement it as we are welcoming contributors to open-source software and get in contact with us, we happily answer requests for collaboration with attached CV and your programming experience!