Add liveness probe to fluentd

This adds a liveness probe to the fluentd chart. This probe will
simply perform a tcpSocket check on the same port the readiness
probe executes the check on.

Change-Id: I768b23d36d50d6f6938f5588bea71e97aeb624b9

fluent-logging/templates/deployment-fluentd.yaml

@@ -117,11 +117,18 @@ spec:
               containerPort: {{ tuple "fluentd" "internal" "service" . | include "helm-toolkit.endpoints.endpoint_port_lookup" }}
             - name: metrics
               containerPort: {{ tuple "fluentd" "internal" "metrics" . | include "helm-toolkit.endpoints.endpoint_port_lookup" }}
+            - name: liveness
+              containerPort: 9880
           readinessProbe:
             tcpSocket:
               port: {{ tuple "fluentd" "internal" "service" . | include "helm-toolkit.endpoints.endpoint_port_lookup" }}
             initialDelaySeconds: 20
             periodSeconds: 10
+          livenessProbe:
+            tcpSocket:
+              port: {{ tuple "fluentd" "internal" "service" . | include "helm-toolkit.endpoints.endpoint_port_lookup" }}
+            initialDelaySeconds: 60
+            timeoutSeconds: 10
           env:
             - name: FLUENTD_PORT
               value: {{ tuple "fluentd" "internal" "service" . | include "helm-toolkit.endpoints.endpoint_port_lookup" | quote }}

fluent-logging/values.yaml

@@ -237,15 +237,24 @@ conf:
         type: monitor_agent
         bind: 0.0.0.0
         port: 24220
+    - http_health_checking:
+        header: source
+        type: http
+        bind: 0.0.0.0
+        port: 9880
     - fluentbit_forward:
         header: source
         type: forward
         port: "#{ENV['FLUENTD_PORT']}"
         bind: 0.0.0.0
-    - filter_fluentd_logs:
+    - filter_fluentbit_logs:
         header: match
         expression: "fluent.**"
         type: "null"
+    - filter_fluentd_logs:
+        header: match
+        expression: "fluentd.**"
+        type: "null"
       # NOTE(srwilkers): Look for specific keywords in the log key to determine
       # log level of event
     - tag_kubernetes_log_level:
@@ -548,6 +557,8 @@ endpoints:
         default: 24224
       metrics:
         default: 24220
+      liveness:
+        default: 9880
   prometheus_fluentd_exporter:
     namespace: null
     hosts: