Hi community. Am deploying to EKS and have just started seeing an error when applying prerequisite helm chart:

kubectl get pods -n datahub
NAME                                               READY   STATUS    RESTARTS   AGE
elasticsearch-master-0                             0/1     Pending   0          16m
elasticsearch-master-1                             0/1     Pending   0          16m
elasticsearch-master-2                             0/1     Pending   0          16m
prerequisites-cp-schema-registry-cf79bfccf-zxp2d   1/2     Error     7          16m
prerequisites-kafka-0                              0/1     Pending   0          16m
prerequisites-zookeeper-0                          0/1     Pending   0          16m

Error coming from cp-schema-registry pod:

kubectl logs -n datahub -f pod/prerequisites-cp-schema-registry-cf79bfccf-zxp2d
error: a container name must be specified for pod prerequisites-cp-schema-registry-cf79bfccf-zxp2d, choose one of: [prometheus-jmx-exporter cp-schema-registry-server]

Has anyone seen this before?

Yes, we've seen this before. If you accidentally deploy the prerequisites chart without the "scaled down" elastic dependencies before adding them, the "standard" storage class won't be created and it will be for ever pending. The solution is to change it to "hostpath".

@millions-jelly-76272 how many nodes are in your EKS cluster and how much resources are you giving the nodes? Also can you run

kubectl describe pod <<pod-name>>

for the pending pods and post the results? With the recommended EKS setup (https://datahubproject.io/docs/deploy/aws#start-up-a-kubernetes-cluster-on-aws-eks) with at least 3 EC2 m3.large nodes, it should be able to bring up the pods without issues

Thanks guys. Issue goes away after I increased size of the nodes. Cheers 🙂

awesome!!