feat: deprecate DYN_SYSTEM_ENABLED in favor of DYN_SYSTEM_PORT (#4082)

Signed-off-by: Keiven Chang <[email protected]>
Co-authored-by: Keiven Chang <[email protected]>

Author: keivenchang

components/src/dynamo/frontend/main.py

@@ -251,6 +251,16 @@ async def async_main():
     dump_config(flags.dump_config_to, flags)
     is_static = bool(flags.static_endpoint)  # true if the string has a value
 
+    # Warn if DYN_SYSTEM_PORT is set (frontend doesn't use system metrics server)
+    if os.environ.get("DYN_SYSTEM_PORT"):
+        logger.warning(
+            "=" * 80 + "\n"
+            "WARNING: DYN_SYSTEM_PORT is set but NOT used by the frontend!\n"
+            "The frontend does not expose a system metrics server.\n"
+            "Only backend workers should set DYN_SYSTEM_PORT.\n"
+            "Use --http-port to configure the frontend HTTP API port.\n" + "=" * 80
+        )
+
     # Configure Dynamo frontend HTTP service metrics prefix
     if flags.metrics_prefix is not None:
         prefix = flags.metrics_prefix.strip()

container/README.md

@@ -357,13 +357,13 @@ python -m dynamo.frontend &
 
 # 4. Start worker backend (choose one framework):
 # vLLM
-DYN_SYSTEM_ENABLED=true DYN_SYSTEM_PORT=8081 python -m dynamo.vllm --model Qwen/Qwen3-0.6B --gpu-memory-utilization 0.20 --enforce-eager --no-enable-prefix-caching --max-num-seqs 64 &
+DYN_SYSTEM_PORT=8081 python -m dynamo.vllm --model Qwen/Qwen3-0.6B --gpu-memory-utilization 0.20 --enforce-eager --no-enable-prefix-caching --max-num-seqs 64 &
 
 # SGLang
-DYN_SYSTEM_ENABLED=true DYN_SYSTEM_PORT=8081 python -m dynamo.sglang --model Qwen/Qwen3-0.6B --mem-fraction-static 0.20 --max-running-requests 64 &
+DYN_SYSTEM_PORT=8081 python -m dynamo.sglang --model Qwen/Qwen3-0.6B --mem-fraction-static 0.20 --max-running-requests 64 &
 
 # TensorRT-LLM
-DYN_SYSTEM_ENABLED=true DYN_SYSTEM_PORT=8081 python -m dynamo.trtllm --model Qwen/Qwen3-0.6B --free-gpu-memory-fraction 0.20 --max-num-tokens 8192 --max-batch-size 64 &
+DYN_SYSTEM_PORT=8081 python -m dynamo.trtllm --model Qwen/Qwen3-0.6B --free-gpu-memory-fraction 0.20 --max-num-tokens 8192 --max-batch-size 64 &
 ```
 
 **Framework-Specific GPU Memory Arguments:**

deploy/cloud/operator/docs/footer.md

@@ -135,9 +135,8 @@ The operator automatically injects environment variables based on component type
 
 ### Worker Components
 
-- **`DYN_SYSTEM_ENABLED`**: `true`
+- **`DYN_SYSTEM_PORT`**: `9090` (automatically enables the system metrics server)
 - **`DYN_SYSTEM_USE_ENDPOINT_HEALTH_STATUS`**: `["generate"]`
-- **`DYN_SYSTEM_PORT`**: `9090`
 
 ### Planner Components
 
@@ -232,4 +231,4 @@ For users who want to understand the implementation details or contribute to the
 - All these defaults can be overridden by explicitly specifying values in your DynamoComponentDeployment or DynamoGraphDeployment resources
 - User-specified probes (via `livenessProbe`, `readinessProbe`, or `startupProbe` fields) take precedence over operator defaults
 - For multinode deployments, some defaults are modified or removed as described above to accommodate distributed execution patterns
-- The `extraPodSpec.mainContainer` field can be used to override probe configurations set by the operator
+- The `extraPodSpec.mainContainer` field can be used to override probe configurations set by the operator

deploy/cloud/operator/internal/controller/dynamocomponentdeployment_controller_test.go

@@ -827,7 +827,6 @@ func TestDynamoComponentDeploymentReconciler_generateLeaderWorkerSet(t *testing.
 											{Name: "DYN_NAMESPACE", Value: "default"},
 											{Name: "DYN_PARENT_DGD_K8S_NAME", Value: "test-lws-deploy"},
 											{Name: "DYN_PARENT_DGD_K8S_NAMESPACE", Value: "default"},
-											{Name: "DYN_SYSTEM_ENABLED", Value: "true"},
 											{Name: "DYN_SYSTEM_PORT", Value: "9090"},
 											{Name: "DYN_SYSTEM_USE_ENDPOINT_HEALTH_STATUS", Value: "[\"generate\"]"},
 											{Name: "TEST_ENV_FROM_DYNAMO_COMPONENT_DEPLOYMENT_SPEC", Value: "test_value_from_dynamo_component_deployment_spec"},
@@ -941,7 +940,6 @@ func TestDynamoComponentDeploymentReconciler_generateLeaderWorkerSet(t *testing.
 											{Name: "DYN_NAMESPACE", Value: "default"},
 											{Name: "DYN_PARENT_DGD_K8S_NAME", Value: "test-lws-deploy"},
 											{Name: "DYN_PARENT_DGD_K8S_NAMESPACE", Value: "default"},
-											{Name: "DYN_SYSTEM_ENABLED", Value: "true"},
 											{Name: "DYN_SYSTEM_PORT", Value: "9090"},
 											{Name: "DYN_SYSTEM_USE_ENDPOINT_HEALTH_STATUS", Value: "[\"generate\"]"},
 											{Name: "TEST_ENV_FROM_DYNAMO_COMPONENT_DEPLOYMENT_SPEC", Value: "test_value_from_dynamo_component_deployment_spec"},

deploy/cloud/operator/internal/dynamo/component_worker.go

@@ -74,10 +74,6 @@ func (w *WorkerDefaults) GetBaseContainer(context ComponentContext) (corev1.Cont
 	}
 
 	container.Env = append(container.Env, []corev1.EnvVar{
-		{
-			Name:  "DYN_SYSTEM_ENABLED",
-			Value: "true",
-		},
 		{
 			Name:  "DYN_SYSTEM_USE_ENDPOINT_HEALTH_STATUS",
 			Value: "[\"generate\"]",

deploy/cloud/operator/internal/dynamo/graph_test.go

@@ -1818,10 +1818,6 @@ func TestGenerateGrovePodCliqueSet(t *testing.T) {
 														Name:  "DYNAMO_POD_GANG_SET_REPLICAS",
 														Value: "1",
 													},
-													{
-														Name:  "DYN_SYSTEM_ENABLED",
-														Value: "true",
-													},
 													{
 														Name:  "DYN_SYSTEM_PORT",
 														Value: "9090",
@@ -1971,10 +1967,6 @@ func TestGenerateGrovePodCliqueSet(t *testing.T) {
 														Name:  "DYNAMO_POD_GANG_SET_REPLICAS",
 														Value: "1",
 													},
-													{
-														Name:  "DYN_SYSTEM_ENABLED",
-														Value: "true",
-													},
 													{
 														Name:  "DYN_SYSTEM_PORT",
 														Value: "9090",
@@ -2623,10 +2615,6 @@ func TestGenerateGrovePodCliqueSet(t *testing.T) {
 														Name:  "DYNAMO_POD_GANG_SET_REPLICAS",
 														Value: "1",
 													},
-													{
-														Name:  "DYN_SYSTEM_ENABLED",
-														Value: "true",
-													},
 													{
 														Name:  "DYN_SYSTEM_PORT",
 														Value: "9090",
@@ -2763,10 +2751,6 @@ func TestGenerateGrovePodCliqueSet(t *testing.T) {
 														Name:  "DYNAMO_POD_GANG_SET_REPLICAS",
 														Value: "1",
 													},
-													{
-														Name:  "DYN_SYSTEM_ENABLED",
-														Value: "true",
-													},
 													{
 														Name:  "DYN_SYSTEM_PORT",
 														Value: "9090",
@@ -4595,7 +4579,6 @@ func TestGenerateBasePodSpec_Worker(t *testing.T) {
 							{Name: "DYN_NAMESPACE", Value: ""},
 							{Name: "DYN_PARENT_DGD_K8S_NAME", Value: "test-deployment"},
 							{Name: "DYN_PARENT_DGD_K8S_NAMESPACE", Value: "default"},
-							{Name: "DYN_SYSTEM_ENABLED", Value: "true"},
 							{Name: "DYN_SYSTEM_PORT", Value: "9090"},
 							{Name: "DYN_SYSTEM_USE_ENDPOINT_HEALTH_STATUS", Value: "[\"generate\"]"},
 						},

deploy/helm/chart/templates/deployment.yaml

@@ -113,8 +113,6 @@ spec:
         - name: DYN_HTTP_PORT
           value: "{{ $.Values.dynamoPort | default 8000 }}"
         {{- else if $serviceSpec.componentType | eq "worker" }}
-        - name: DYN_SYSTEM_ENABLED
-          value: "true"
         - name: DYN_SYSTEM_PORT
           value: "{{ $.Values.dynamoSystemPort | default 9090 }}"
         - name: DYN_SYSTEM_USE_ENDPOINT_HEALTH_STATUS

deploy/helm/chart/templates/grove-podgangset.yaml

@@ -105,8 +105,6 @@ spec:
             - name: DYN_HTTP_PORT
               value: "{{ $.Values.dynamoPort | default 8000 }}"
             {{- else if $serviceSpec.componentType | eq "worker" }}
-            - name: DYN_SYSTEM_ENABLED
-              value: "true"
             - name: DYN_SYSTEM_PORT
               value: "{{ $.Values.dynamoSystemPort | default 9090 }}"
             - name: DYN_SYSTEM_USE_ENDPOINT_HEALTH_STATUS

deploy/metrics/prometheus.yml

@@ -41,7 +41,7 @@ scrape_configs:
     static_configs:
       - targets: ['host.docker.internal:8000']  # on the "monitoring" network
 
-  # Launch via: DYN_SYSTEM_ENABLED=true DYN_SYSTEM_PORT=8081 dynamo.<backend> ...
+  # Launch via: DYN_SYSTEM_PORT=8081 dynamo.<backend> ...
   # If you want to update the scrape_interval, you may want to also update component.rs's MAX_DELAY
   - job_name: 'dynamo-backend'
     scrape_interval: 6s

docs/backends/sglang/prometheus.md

@@ -43,8 +43,7 @@ To see the actual metrics available in your SGLang version:
 ### 1. Launch SGLang with Metrics Enabled
 
 ```bash
-# Set environment variables
-export DYN_SYSTEM_ENABLED=true
+# Set system metrics port (automatically enables metrics server)
 export DYN_SYSTEM_PORT=8081
 
 # Start SGLang worker with metrics enabled