https://www.gooddata.com/ logo
Join Slack
Powered by
<@U021NRNA866> I am trying to install good-data-cn...
# gooddata-cn
v
@Robert Moucha I am trying to install good-data-cn on AWS/EKS according to the helm installation instructions here: https://www.gooddata.com/developers/cloud-native/doc/2.3/deploy-and-install/cloud-native/helm-chart-installation/ I'm on a 3x nodes with c6a.xlarge instance types, but the pulsar helm installation never seems to get the desired number of nodes in their stateful set... I am seeing only a single zookeeper-0 pod and not 3x does this need to have 3x replicas explicitly specified... I am using the 2.9.2 version of the helm. It seems like it's waiting on zookeeper to have 3 replica instances? It looks like it's ignoring the default replica instances in the pulsar chart... is there anything else I need to do to get the pulsar chart to load?
Copy code
➜  good-data-helm git:(develop) ✗ make pulsar
kubectl apply -k ./
namespace/gooddata-cn unchanged
namespace/ingress-nginx unchanged
namespace/pulsar unchanged
secret/gooddata-cn-license configured
helm upgrade --install --namespace pulsar --version 2.9.2 \
    -f ./customized-values-pulsar.yaml \
    pulsar apache/pulsar
Release "pulsar" does not exist. Installing it now.
W0323 08:44:17.745100   45176 warnings.go:70] policy/v1beta1 PodDisruptionBudget is deprecated in v1.21+, unavailable in v1.25+; use policy/v1 PodDisruptionBudget
W0323 08:44:17.804849   45176 warnings.go:70] policy/v1beta1 PodDisruptionBudget is deprecated in v1.21+, unavailable in v1.25+; use policy/v1 PodDisruptionBudget
W0323 08:44:17.859984   45176 warnings.go:70] policy/v1beta1 PodDisruptionBudget is deprecated in v1.21+, unavailable in v1.25+; use policy/v1 PodDisruptionBudget
W0323 08:44:19.189758   45176 warnings.go:70] policy/v1beta1 PodDisruptionBudget is deprecated in v1.21+, unavailable in v1.25+; use policy/v1 PodDisruptionBudget
W0323 08:44:19.190324   45176 warnings.go:70] policy/v1beta1 PodDisruptionBudget is deprecated in v1.21+, unavailable in v1.25+; use policy/v1 PodDisruptionBudget
W0323 08:44:19.190921   45176 warnings.go:70] policy/v1beta1 PodDisruptionBudget is deprecated in v1.21+, unavailable in v1.25+; use policy/v1 PodDisruptionBudget
NAME: pulsar
LAST DEPLOYED: Thu Mar 23 08:44:17 2023
NAMESPACE: pulsar
STATUS: deployed
REVISION: 1
TEST SUITE: None
customized-values-pulsar.yaml
I'm getting these messages in the pulsar-init pod:
Copy code
➜  ~ kubectl logs -c wait-zookeeper-ready pulsar-pulsar-init-6h96x -n pulsar
Server:		172.20.0.10
Address:	172.20.0.10#53

** server can't find pulsar-zookeeper-2.pulsar-zookeeper.pulsar: NXDOMAIN
I'm trying to add:
Copy code
zookeeper:
  replicaCount: 3
To the pulsar values... but it doesn't look like it's having an effect on the number of zookeeper replicas?
r
it may take some time until ZK cluster is established. StatefulSet replicas are deployed sequentially from 0 to 2.
v
ok
is there a pod I would look at for status?
r
Pulsar-init pod polls until all ZK pods are ready. Check pods pulsar-zookeeper-{0,1,2}
v
no log messages from the pulsar-zookeeper-0 pod yet
8 minutes have gone by
r
what is the pod status?
v
pending
r
aha
v
no events
r
I guess it is waiting for disk to be attached
v
ok
it's that gp2 default storage class in eks
and no volume
r
what eks version do you use? 1.24?
v
yessir
image.png
r
fine. Check EKS addons and install EBS CSI driver. It should help
Newer k8s dropped so-called in-tree drivers and they need to be installed separately
v
yup you are right: 
led to provision volume with StorageClass "gp2": rpc error: code = Internal desc = Could not create volume "pvc-d56ca03b-bfb6-4196-b7fd-09b7e8009d89": could not create volume in EC2: UnauthorizedOperation: You are not authorized to perform this operation.
let me go figure that out
hehe you my friend are a legend!
image.png
Not quite there yet... but it is moving a little
image.png
r
great, pulsar seems to be operation.
Open in Slack
PreviousNext
Powered by