The Container Service Dashboard displays the results of our automated testing suite.
Each task is tested once per hour on multiple Kubernetes versions.
The "Task History" panels show the previous results for each task at each hour. Green means that the tests on all Kubernetes versions passed in that hour. Red means that all tests failed, and orange means that some passed and some failed.
The lower table shows what percent of tests passed on each Kubernetes version, over the entire selected time period.
The tests ensure that the listed functionality is working as expected (with the relevant labels set at cluster creation).
|k8s-cephfs||Creates PVCs for cephfs shares with reclaim policies reclaim and delete, and checks that they are created/deleted as expected.|
|k8s-cern-enabled||Creates a cluster with the cern_enabled label set, and ensures that grid certificates are created on the nodes.|
|k8s-cinder||Provisions a cinder volume using a PVC and mounts it to a pod.|
|k8s-cvmfs||Creates a pod with two containers, checking that the CVMFS volumes for ATLAS and CMS are properly mounted.|
|k8s-dns||Checks that DNS is working both for cluster internal service names and for an external URL (openstack.cern.ch).|
|k8s-eos||Checks that /eos is mounted onto the node host filesystem and all user directories a-z are listed.|
|Sets up one insecure ingress and one TLS terminated ingress for a service and checks that both are reachable.|
|k8s-landb-sync||Creates an ingress in the cluster and ensures that landb-sync sets the correct alias on the worker node.|
|k8s-logging||Ensures that user and system logs are collected and pushed to CERN central logging.|
|k8s-monitoring||Checks that node metrics are available, and that prometheus/grafana are up and running.|
|k8s-service-type-lb||Creates a LoadBalancer type service, checks that it receives an external IP, and that a curl to the IP reaches the service.|
|swarm||Creates a Swarm cluster and checks that an nginx container can be successfully created.|