Pat Ledgerwood
Pat Ledgerwood
- [ ] Kibana dashboard for cluster wide Notebook usage - [ ] Granfana dashboard for cluster wide user resource usage (notebook pods and ml pods) - [ ] Grafana...
#### Describe the bug The version of code-server that we bundle the juypterlab docker image has a bug where the user settings are stored only in the browser and not...
- [ ] How are these manifests deployed into dev/prod (i.e. what is the general workflow to update manifests and make releases into dev/prod)? - [ ] How are these...
The BLOB CSI implementation shares large (10TB) PV's with all user namespaces but these are presented as individual PVC's and are being overcounted by Kubecost, they need to be ignored...
Shared namespaces created manually by the AAW team need aggregation labels for kubecost, these will need to be provided by the user for the namespace creation and should be added...
Namespaces which host platform services used by all users need to be identified by kubecost so the costs can be shared with user/project namespaces. https://kubecost.aaw.cloud.statcan.ca/settings.html Current list: `aad-pod-identity-system,boathouse-system,cert-manager-system,daaas-system,gatekeeper-system,goofys-injector-system,istio-operator-system,istio-system,jfrog-system,kube-system,kubecost-system,minio-opa-system,minio-operator-system,minio-premium-system,minio-standard-system,prometheus-system,statcan-system,vault-agent-system,velero-system` We should...
Namespaces created by individual users need aggregation labels for kubecost, ideally these could be identified on creation (extracted from active directory?). Support in kubecost for these labels already exists: https://gitlab.k8s.cloud.statcan.ca/cloudnative/terraform/modules/terraform-statcan-kubernetes-core-platform/-/blob/main/kubecost.tf...
Integrate APM metrics into user facing (web) apps in AAW for better performance and usage statistics. - [ ] Kubeflow - https://github.com/StatCan/kubeflow/blob/stc-master/components/centraldashboard/public/index.html (client side) - https://www.elastic.co/guide/en/apm/agent/rum-js/current/install-the-agent.html (client side) - https://www.elastic.co/guide/en/apm/agent/nodejs/current/typescript.html...
- [ ] Grafana dashboard with namespace as variable to display user pod metrics, should be embeddable in size - [ ] Metrics should then be embedded into kubeflow with...
Elasticserch index retention must be appropriate for the size of the disk and the data being retained. - [x] Update elasticsearch / kibana / heartbeat to latest version (dev +...