12/05/2019, 7:16 AM
Hi I have one stack up and running with 2 k8 clusters with multiple helm charts deployed it. After the above stack is up, I tried to "pulumi up" on another different stack "in parallel" with same above configurations of helm charts, some of the charts are failing : "error: Plan apply failed: 2 errors occurred: * resource kibana-kibana was successfully created, but the Kubernetes API server reported that it failed to fully initialize or become live: 'kibana-kibana' timed out waiting to be Ready * Service was not allocated an IP address; does your cloud provider support this?" any idea on what is happening?


12/05/2019, 9:32 AM
Seams than your k8s cni can't allocate new one ip for your clusterIp service
Check manually is it possible to create the same service in k8s cluster
Also check your cidr prefixs for the clusterip network, is it anough ips for all your services in the cluster


12/12/2019, 12:50 AM
thank you very much. I have very less knowledge on cni.... I am using default vpc and my code for single stack is as simple as this: cluster = new eks.Cluster("datalake-cluster", { // desiredCapacity: 7, //minSize: 5, //maxSize: 7, desiredCapacity: 5, minSize: 5, maxSize: 5, instanceType: "t2.large", deployDashboard: false, }, { customTimeouts: { create: "2m", delete: "2m", update: "2m" } } and same for 2nd cluster in the same stack... When you say, check manually, do you mean creating eks using aws UI?
👍 1
figured out there was a limit of number of load balancers per region. we requested for increase and it is working fine now. do you see opportunity for better error handling on the pulumi side @quaint-garden-96746?