Can kong api gateway communicate to cassandra cluster with in k3s

We have k3s cluster with postgres as datastore, also we have cassandra cluster deployed in k3s. We have deployed kong ingress with cassandra db using https://github.com/Kong/kubernetes-ingress-controller/blob/main/deploy/single/all-in-one-postgres.yaml
replaced postgres with cassandra as below in yaml file.

  • name: KONG_DATABASE
    value: cassandra
    - name: KONG_CASSANDRA_HOST
    value: cassandra
    - name: KONG_CASSANDRA_PASSWORD

kong_cassandra_host means system ip where cassandra installed or username ? since cassandra is deployed as cluster on 2 nodes so which node ip to provide…
there is no sample yaml file for kong with cassandra…

deployment.apps/ingress-kong and statefulset.apps/cassandra are not running due to which pod/kong-migrations and pod/ingress-kong pods are not running.
attached screenshot…

[root@kong ~]# kubectl get logs pod/ingress-kong-844c848f94-7ml6q -n kong
error: there is no need to specify a resource type as a separate argument when passing arguments in resource/name form (e.g. ‘kubectl get resource/<resource_name>’ instead of ‘kubectl get resource resource/<resource_name>’

Receiving 503 service unavailable error from accessing kong admin api…

curl -i $PROXY_IP

Support for Cassandra-backed controller instances is deprecated and will be removed entirely in the near future. You should switch to either a Postgres-backed or DB-less instance: there shouldn’t be any reason to use Cassandra with the controller, as the source of truth for controller-managed Kong clusters is the Kubernetes configuration in etcd, and etcd provides similar distributed database functionality.

For what it’s worth, KONG_CASSANDRA_HOST can be any Cassandra instance in the Cassandra cluster, as the Cassandra driver only uses it to find a single initial point of contact before using Cassandra’s native cluster member discovery mechanism. Using cassandra as you’ve done will use Kubernetes service discovery to return a list of all active Cassandra Pods: if none are ready, as shown in your screenshot, that list will be empty. You’d need to fix whatever issues are preventing the Cassandra instances from coming online for it to work, though again, we recommend against using Cassandra at all since you’ll have to migrate away from it anyway in the future.