test: skip persistent nightly CI failures and provide associated tickets (#8006)

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

tests/basic/test_autodeploy_backend.py

@@ -166,6 +166,7 @@ def send_completion_request(
         raise
 
 
+@pytest.mark.skip(reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2607")
 @pytest.mark.trtllm
 @pytest.mark.e2e
 @pytest.mark.slow

tests/fault_tolerance/cancellation/test_vllm.py

@@ -296,6 +296,7 @@ def test_request_cancellation_vllm_aggregated(
                 )
 
 
+@pytest.mark.skip(reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2606")
 @pytest.mark.timeout(150)  # 3x average
 @pytest.mark.nightly
 @pytest.mark.gpu_2

tests/gms/integration/test_external_weight_mgr.py

@@ -261,6 +261,7 @@ def test_external_weight_mgr_vllm(
     )
 
 
+@pytest.mark.skip(reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2567")
 @pytest.mark.sglang
 @pytest.mark.e2e
 @pytest.mark.gpu_1

tests/gms/integration/test_gms_shadow_failover.py

@@ -465,6 +465,7 @@ def test_gms_shadow_engine_failover_vllm(
     )
 
 
+@pytest.mark.skip(reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2567")
 @pytest.mark.sglang
 @pytest.mark.e2e
 @pytest.mark.gpu_1

tests/gms/integration/test_gms_sleep_wake.py

@@ -208,6 +208,7 @@ def test_gms_basic_sleep_wake_vllm(
     )
 
 
+@pytest.mark.skip(reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2567")
 @pytest.mark.sglang
 @pytest.mark.e2e
 @pytest.mark.gpu_1

tests/router/test_router_e2e_with_sglang.py

@@ -318,6 +318,7 @@ def test_router_decisions_sglang_dp(
     )
 
 
+@pytest.mark.skip(reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2603")
 @pytest.mark.gpu_2
 @pytest.mark.nightly
 @pytest.mark.parametrize("request_plane", ["nats"], indirect=True)

tests/router/test_router_e2e_with_trtllm.py

@@ -302,6 +302,7 @@ def test_router_decisions_trtllm_multiple_workers(
     )
 
 
+@pytest.mark.skip(reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2609")
 @pytest.mark.gpu_2
 @pytest.mark.nightly
 @pytest.mark.parametrize("request_plane", ["nats"], indirect=True)

tests/serve/multimodal_profiles/vllm.py

@@ -84,7 +84,12 @@ VLLM_MULTIMODAL_PROFILES: list[MultimodalModelProfile] = [
         short_name="qwen2-audio-7b",
         topologies={
             "agg": TopologyConfig(
-                marks=[pytest.mark.post_merge],
+                marks=[
+                    pytest.mark.skip(
+                        reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2604"
+                    ),
+                    pytest.mark.post_merge,
+                ],
                 timeout_s=600,
                 env={"DYN_CHAT_PROCESSOR": "vllm"},
             ),

tests/serve/test_sglang.py

@@ -271,6 +271,9 @@ sglang_configs = {
         directory=sglang_dir,
         script_name="agg.sh",
         marks=[
+            pytest.mark.skip(
+                reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2602"
+            ),
             pytest.mark.gpu_1,
             pytest.mark.profiled_vram_gib(
                 19.1

tests/serve/test_trtllm.py

@@ -239,6 +239,9 @@ trtllm_configs = {
         directory=trtllm_dir,
         script_name="agg_multimodal_router.sh",
         marks=[
+            pytest.mark.skip(
+                reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2608"
+            ),
             pytest.mark.gpu_1,
             pytest.mark.trtllm,
             pytest.mark.multimodal,

tests/serve/test_vllm.py

@@ -423,6 +423,9 @@ vllm_configs = {
         directory=vllm_dir,
         script_name="agg_multimodal.sh",
         marks=[
+            pytest.mark.skip(
+                reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2604"
+            ),
             pytest.mark.gpu_1,  # agg_multimodal.sh uses single GPU
             pytest.mark.multimodal,
             pytest.mark.nightly,
@@ -626,6 +629,7 @@ def test_serve_deployment(
     run_serve_deployment(config, request, ports=dynamo_dynamic_ports)
 
 
+@pytest.mark.skip(reason="Nightly CI failure: https://linear.app/nvidia/issue/DYN-2605")
 @pytest.mark.vllm
 @pytest.mark.e2e
 @pytest.mark.gpu_2