Post Install Testing

Pod Health

The initial step post deployment is to check the health of pods. This can be done using Lens or antoher client or the kubectl command line. After the script has finished Kubernetes will be busy trying to get the cluster into the state specified in the generated yaml that was applied.

Zookeeper and Kafka

Zookeeper is used by Kafka to store metadata and Kafka is the messaging backbone of the new fraXses platform. These are the two most important stateful sets. They should come up slowly in order. Whilst they are coming up, most of the other pods will be in an indeterminate state because they can't yet communicate with Kafka.

Pods with Persistent Volumes

The second important group are those deployments with persistent volume claims. THese include:

Deployment NameComponentComment
fraxses-mdbMeta DatabaseHolds all fraXses metadata
fraxses-csvCompute ServerModified Apache Spark
fraxses-elsElasticsearchStore for log data
fraxses-legLegozLegoz UI (this also requires connection to the gateway api )
fraxses-pgrPostgressLegoz UI (this also requires connection to the gateway api )

Node Port Access

Once all pods are green and healthy, you should check the access to the node ports. The node ports can be

ComponentNode Port
fraxses-mdbMeta Database
fraxses-csvCompute Server

Gateway API access