[helm] Use grpc-based liveness probe

The previous commit introduced a gRPC server that can be used for
the liveness probe. This patch changes helm to make that default instead
of the tetra status based liveness probe.

The user can still use the tetra status based liveness probe by
defining a values file similar to:
tetragon:
  livenessProbe:
     timeoutSeconds: 60
     exec:
       command:
       - tetra
       - status
       - --server-address
       - "54321"
       - --retries
       - "5"

Signed-off-by: Anastasios Papagiannis <[email protected]>

install/kubernetes/tetragon/templates/_container_tetragon.tpl

@@ -67,17 +67,12 @@
 {{- if .Values.tetragon.livenessProbe }}
   livenessProbe:
   {{- toYaml .Values.tetragon.livenessProbe | nindent 4 }}
-{{- else if .Values.tetragon.grpc.enabled }}
+{{- else if .Values.tetragon.healthGrpc.enabled }}
   livenessProbe:
      timeoutSeconds: 60
-     exec:
-       command:
-       - tetra
-       - status
-       - --server-address
-       - {{ .Values.tetragon.grpc.address }}
-       - --retries
-       - "5"
+     grpc:
+      port: {{ .Values.tetragon.healthGrpc.port }}
+      service: "liveness"
 {{- end -}}
 {{- end -}}
 

kind-config.yaml

@@ -0,0 +1,7 @@
+apiVersion: kind.x-k8s.io/v1alpha4
+kind: Cluster
+nodes:
+  - role: control-plane
+    extraMounts:
+      - hostPath: /proc
+        containerPath: /procHost

values.yaml

@@ -0,0 +1,21 @@
+tetragon:
+  hostProcPath: "/procHost"
+  healthGrpc:
+    enabled: true
+    port: 6780
+    interval: 11
+  # livenessProbe:
+  #    timeoutSeconds: 60
+  #    exec:
+  #      command:
+  #      - tetra
+  #      - status
+  #      - --server-address
+  #      - "54321"
+  #      - --retries
+  #      - "5"
+  image:
+    override: "quay.io/cilium/tetragon-ci:e3692fc57dd830b999fbbbd4019672a0039fd0ac"
+tetragonOperator:
+  image:
+    override: "quay.io/cilium/tetragon-operator-ci:e3692fc57dd830b999fbbbd4019672a0039fd0ac"