chore: remove old flag

Author: mchmarny

distros/kubernetes/nvsentinel/charts/syslog-health-monitor/templates/configmap.yaml

@@ -31,10 +31,11 @@ data:
   # Previously, kata-specific journal tags (e.g., "-u containerd.service") were
   # conditionally added here, but these are no longer needed because:
   # 1. Kata detection happens at the deployment level via node labels
-  # 2. Each DaemonSet variant (regular/kata) handles journal access appropriately
-  #    through its own hostPath mounts and container configuration
-  # 3. The kata-enabled flag is passed as a command-line argument to configure
-  #    runtime behavior without requiring different ConfigMap content
+  # 2. Each DaemonSet variant (regular/kata) handles journal access differently:
+  #    - Regular: mounts /var/log (file-based logs)
+  #    - Kata: mounts /run/log/journal and /var/log/journal (systemd journal)
+  # 3. The kata variant accesses the systemd journal directly, so it can read
+  #    containerd logs without needing special tags or filters in the ConfigMap
   log_check_definitions.yaml: |
     checks:
       - name: "SysLogsXIDError"

health-monitors/syslog-health-monitor/main.go

@@ -63,8 +63,6 @@ var (
 	metricsPort         = flag.String("metrics-port", "2112", "Port to expose Prometheus metrics on")
 	xidAnalyserEndpoint = flag.String("xid-analyser-endpoint", "",
 		"Endpoint to the XID analyser service.")
-	kataEnabled = flag.String("kata-enabled", "false",
-		"Indicates if this monitor is running in Kata Containers mode (set by DaemonSet variant).")
 )
 
 // ConfigFile matches the top-level structure of the YAML config file
@@ -92,7 +90,7 @@ func run() error {
 		return fmt.Errorf("NODE_NAME env not set and --node-name flag not provided, cannot run")
 	}
 
-	slog.Info("Configuration", "node", nodeName, "kata-enabled", *kataEnabled)
+	slog.Info("Configuration", "node", nodeName)
 
 	// Root context canceled on SIGINT/SIGTERM so goroutines can exit cleanly.
 	root := context.Background()