search

skoussou / servicemesh-playground

RH ServiceMesh setups
17 stars 13 forks source link

Federation Example: Step 1 Failing (RHPS/AWS/RHACM controlled setup) #1

Open karstengresch opened 2 years ago

karstengresch commented 2 years ago

Hi Stelios,

thanks first for this project! So many wonderful scenarios and examples in your GH project, I'm extremely expressed.

Regarding the federation example, I stumbled on this problem (tokens given below are not relevant as the environment is already deleted).

Setup was: RHACM cluster ("OCP 4 ACM Hub" on RHPDS - OCP 4.9 w/ RHACM 2.6.1, "as-is") AWS Cluster 01: "prod03" (OCP 4.11.4) AWS Cluster 02: "prod06" (OCP 4.11.4)

I logged in as kubeadmin (no extra user with cluster-admin permissions created as I assume, kubeadmin always has sufficient privileges) and applied 0-operator-subscription.yaml on each target cluster.

Then I ran ./0-execute-federation-setup-AWS-LB.sh (detailed output in the logfile attached).

I received the following errors in step 1:

 domain: east-mesh.local |oc apply -f -
Error from server (InternalError): error when creating "STDIN": Internal error occurred: failed calling webhook "smcp.mutation.maistra.io": failed to call webhook: Post "https://maistra-admission-controller.openshift-operators.svc:443/mutate-smcp?timeout=10s": dial tcp 10.128.2.33:11999: connect: connection 
refused

and

apiVersion: maistra.io/v1
kind: ServiceMeshMemberRoll
metadata:
  name: default
  namespace: remote-east-mesh-system
spec:
  members:
  - east-travel-agency
  - east-travel-portal
  - east-travel-control |oc apply -f -
Error from server (InternalError): error when creating "STDIN": Internal error occurred: failed calling webhook "smmr.mutation.maistra.io": failed to call webhook: Post "https://maistra-admission-controller.openshift-operators.svc:443/mutate-smmr?timeout=10s": dial tcp 10.128.2.33:11999: connect: connection 
refused

Etc.

Maybe you'd like to have a look into that?

Best wishes,

Karsten

log.txt

karstengresch commented 2 years ago

Update: at least one workaround is to set up the ServiceMesh control planes manually (install operators as per documentation: 0. Elasticsearch (optional) 1. Red Hat Distributed Tracing Platform 2. Red Hat (!) Kiali 3. Red Hat ServiceMesh).

This, at least, let me run the script through (on AWS).

NB: on AWS as per documentation this security spec attribute could be needed 

security:
    identity:
      type: ThirdParty