Eks Fluentbit – ahmadrazalab

In Progress: Setting up EKS with Fluent Bit for Log Aggregation

helm repo add fluent https://fluent.github.io/helm-charts
helm repo update
> values.yaml `https://github.com/fluent/helm-charts/blob/main/charts/fluent-bit/values.yaml`

helm show values  fluent/fluent-bit > tracemypods-values.yaml
helm upgrade --install fluent-bit fluent/fluent-bit --namespace kube-system -f values.yaml


kubectl delete configmap kibana-kibana-helm-scripts -n logging
kubectl delete serviceaccount pre-install-kibana-kibana -n logging
kubectl delete role pre-install-kibana-kibana -n logging
kubectl delete rolebinding pre-install-kibana-kibana -n logging
kubectl delete job pre-install-kibana-kibana -n logging
kubectl delete secret kibana-kibana-es-token -n logging
kubectl delete sa post-delete-kibana-kibana -n logging



kubectl create secret generic elasticsearch-master-certs \
  --from-literal=ca.crt="" \
  -n logging

kubectl create secret generic elasticsearch-master-credentials \
  --from-literal=username=elastic \
  --from-literal=password=changeme \
  -n logging



helm install kibana elastic/kibana -n logging -f kibana-values.yaml --no-hooks

helm upgrade kibana elastic/kibana -n logging -f kibana-values.yaml --no-hooks






## ✅ Corrected and Final Working Flow

### 🔹 Step 1: Create Namespace

```bash
kubectl create namespace logging

🔹 Step 2: Add Helm Repo

helm repo add elastic https://helm.elastic.co
helm repo update

🔹 Step 3: Elasticsearch Deployment

Update your elasticsearch-values.yaml:

clusterName: "elasticsearch"
nodeGroup: "master"
 
secret:
  enabled: true
  password: "changeme"
 
createCert: true

Now deploy:

helm upgrade --install elasticsearch elastic/elasticsearch \
  -n logging -f elasticsearch-values.yaml

Wait until pods are ready:

kubectl get pods -n logging

🔹 Step 4: Verify Secrets

Find the correct names:

kubectl get secrets -n logging

Example output may show:

elasticsearch-master-certs ✅ (not minikube-es-master-certs)
elasticsearch-master-credentials ✅

Then extract and recreate CA secret:

kubectl get secret elasticsearch-master-certs -n logging \
  -o jsonpath="{.data.ca\.crt}" | base64 --decode > ca.crt
 
kubectl create secret generic elasticsearch-certificates \
  --from-file=ca.crt=ca.crt \
  -n logging

🔹 Step 5: `kibana-values.yaml` (Corrected)

elasticsearchHosts: "https://elasticsearch-master.logging.svc:9200"
protocol: https
elasticsearchCertificateSecret: elasticsearch-certificates
elasticsearchCertificateAuthoritiesFile: ca.crt
elasticsearchCredentialSecret: elasticsearch-master-credentials
createElasticsearchToken: false

This disables the failing token creation job and uses your existing Elasticsearch credentials.

🔹 Step 6: Deploy Kibana

helm install kibana elastic/kibana \
  -n logging -f kibana-values.yaml

Then confirm:

kubectl get pods -n logging
kubectl get svc -n logging

✅ Done!

Kibana should now come up without:

The token error
The DNS error (ENOTFOUND elasticsearch-master...)

Prom.yml Eks Grafana Athena S3