Remove redundant --http-port flag from launch scripts

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

Author: keivenchang

examples/backends/sglang/launch/disagg_router.sh

@@ -12,8 +12,8 @@ cleanup() {
 trap cleanup EXIT INT TERM
 
 # run ingress
+# DYN_HTTP_PORT env var is read by dynamo.frontend (defaults to 8000 if not set)
 python3 -m dynamo.frontend \
- --http-port=${DYN_HTTP_PORT:-8000} \
  --router-mode kv \
  --kv-overlap-score-weight 0 \
  --router-reset-states &

examples/backends/trtllm/launch/agg_router.sh

@@ -19,7 +19,8 @@ trap cleanup EXIT INT TERM
 
 
 # run frontend
-python3 -m dynamo.frontend --router-mode kv --http-port ${DYN_HTTP_PORT:-8000} &
+# DYN_HTTP_PORT env var is read by dynamo.frontend (defaults to 8000 if not set)
+python3 -m dynamo.frontend --router-mode kv &
 DYNAMO_PID=$!
 
 # run worker

examples/backends/trtllm/launch/disagg_router.sh

@@ -22,7 +22,8 @@ trap cleanup EXIT INT TERM
 
 
 # run frontend with KV routing for cache-aware optimization
-python3 -m dynamo.frontend --router-mode kv --http-port ${DYN_HTTP_PORT:-8000} &
+# DYN_HTTP_PORT env var is read by dynamo.frontend (defaults to 8000 if not set)
+python3 -m dynamo.frontend --router-mode kv &
 DYNAMO_PID=$!
 
 # run prefill worker

examples/backends/trtllm/launch/gpt_oss_disagg.sh

@@ -14,7 +14,8 @@ trap 'echo Cleaning up...; kill 0' EXIT
 
 
 # run frontend
-python3 -m dynamo.frontend --router-mode round-robin --http-port ${DYN_HTTP_PORT:-8000} &
+# DYN_HTTP_PORT env var is read by dynamo.frontend (defaults to 8000 if not set)
+python3 -m dynamo.frontend --router-mode round-robin &
 
 # With tensor_parallel_size=4, each worker needs 4 GPUs
 # run prefill worker

examples/backends/vllm/launch/agg_kvbm_router.sh

@@ -11,9 +11,9 @@ export PYTHONHASHSEED=0
 MODEL="Qwen/Qwen3-0.6B"
 
 # run frontend + KV router
+# DYN_HTTP_PORT env var is read by dynamo.frontend (defaults to 8000 if not set)
 python -m dynamo.frontend \
     --router-mode kv \
-    --http-port ${DYN_HTTP_PORT:-8000} \
     --router-reset-states &
 
 # run workers with KVBM enabled

examples/backends/vllm/launch/agg_router.sh

@@ -12,9 +12,9 @@ MODEL="Qwen/Qwen3-0.6B"
 BLOCK_SIZE=64
 
 # run frontend + KV router
+# DYN_HTTP_PORT env var is read by dynamo.frontend (defaults to 8000 if not set)
 python -m dynamo.frontend \
     --router-mode kv \
-    --http-port ${DYN_HTTP_PORT:-8000} \
     --router-reset-states &
 
 # run workers

examples/backends/vllm/launch/disagg_kvbm_router.sh

@@ -11,9 +11,9 @@ export PYTHONHASHSEED=0
 MODEL="Qwen/Qwen3-0.6B"
 
 # run decode router with kv-overlap-score-weight 0 for pure load balancing
+# DYN_HTTP_PORT env var is read by dynamo.frontend (defaults to 8000 if not set)
 python -m dynamo.frontend \
     --router-mode kv \
-    --http-port ${DYN_HTTP_PORT:-8000} \
     --kv-overlap-score-weight 0 \
     --router-reset-states &
 

examples/backends/vllm/launch/disagg_router.sh

@@ -13,9 +13,9 @@ BLOCK_SIZE=64
 
 # Start frontend with KV routing
 # The frontend will automatically detect prefill workers and activate an internal prefill router
+# DYN_HTTP_PORT env var is read by dynamo.frontend (defaults to 8000 if not set)
 python -m dynamo.frontend \
     --router-mode kv \
-    --http-port ${DYN_HTTP_PORT:-8000} \
     --router-reset-states &
 
 # two decode workers

tests/serve/launch/template_verifier.sh

@@ -17,7 +17,8 @@ cleanup() {
 trap cleanup EXIT INT TERM
 
 # run ingress
-python3 -m dynamo.frontend --http-port=${DYN_HTTP_PORT:-8000} &
+# DYN_HTTP_PORT env var is read by dynamo.frontend (defaults to 8000 if not set)
+python3 -m dynamo.frontend &
 FRONTEND_PID=$!
 
 # run the mock worker + template validation generate()