Elasticsearch stuck in ApplyingChanges phase on Openshift

[TimTim (Timothy Eichmann)]

Hi!

We're running Openshift 4.11 and we've installed ECK 2.6.2. I start an Elasticsearch instance with this YAML code (copied from the quickstart doc):

cat <<EOF | oc apply -n 008ep-elk -f -
# This sample sets up an Elasticsearch cluster with an OpenShift route
apiVersion: elasticsearch.k8s.elastic.co/v1
kind: Elasticsearch
metadata:
  name: gbg
spec:
  version: 8.6.0
  nodeSets:
  - name: prod
    count: 1
    config:
      node.store.allow_mmap: false
EOF

The pod gets created and is up and running after a while:

$ oc get pods
NAME            READY   STATUS    RESTARTS   AGE
gbg-es-prod-0   1/1     Running   0          17m

Even when I go into the pod and check for health status, I get code green. So everything is working there. But the problem is that the status for Elasticsearch (under Operator) is still saying ApplyingChanges:

$ oc get elasticsearch
NAME   HEALTH    NODES   VERSION   PHASE             AGE
gbg    unknown   1       8.6.0     ApplyingChanges   16m

The question is if this has to do with the Readiness probe that failed for the elasticsearch pod?

[michael.morello]

I've installed ECK 2.6.2 using the operator hub, and then successfully created an Elasticsearch resource using the manifest you provided:

oc get es,pods
NAME                                             HEALTH   NODES   VERSION   PHASE   AGE
elasticsearch.elasticsearch.k8s.elastic.co/gbg   green    1       8.6.0     Ready   73s

NAME                READY   STATUS    RESTARTS   AGE
pod/gbg-es-prod-0   1/1     Running   0          71s

Could you look at the operator's logs to check if there's any error message? Also there may be more information in the Elasticsearch status subresource: oc get elasticsearch.elasticsearch.k8s.elastic.co/gbg --output=jsonpath='{ .status }'