[Fix] Remove MoRI-IO patches from vLLM Disagg benchmarks (#1585)

* fix: remove moriio connector patches after bumping to new nightly vllm image

Signed-off-by: simondanielsson <simon.danielsson99@hotmail.com>

* fix: move read mode envvar to flag

Signed-off-by: simondanielsson <simon.danielsson99@hotmail.com>

* fix: unpin vllm-router image

Signed-off-by: simondanielsson <simon.danielsson99@hotmail.com>

* fix: update to mori_low_latency backend after rename in nightly

Signed-off-by: simondanielsson <simon.danielsson99@hotmail.com>

* fix: pin nightlyies

Signed-off-by: simondanielsson <simon.danielsson99@hotmail.com>

* fix: comments and add perf-changelog.yml

Signed-off-by: simondanielsson <simon.danielsson99@hotmail.com>

* fix: pint router  iamge as well

Signed-off-by: simondanielsson <simon.danielsson99@hotmail.com>

* fix: bump vllm nightly versions

Signed-off-by: simondanielsson <simon.danielsson99@hotmail.com>

* fix: add ulimit nofile

Signed-off-by: simondanielsson <simon.danielsson99@hotmail.com>

---------

Signed-off-by: simondanielsson <simon.danielsson99@hotmail.com>
Co-authored-by: Chun Fang <chun.fang@amd.com>

Author: simondanielsson

benchmarks/multi_node/amd_utils/job.slurm

@@ -317,7 +317,7 @@ export DOCKER_CONT_NAME="container_${ENGINE}_${SANITIZED_USER}_${MODEL_NAME}_${S
 # vLLM external router container.
 # NOTE: vllm/vllm-router only retains ~16 recent nightlies on Docker Hub; older
 # dated tags are garbage-collected (manifest unknown)
-VLLM_ROUTER_IMAGE="${VLLM_ROUTER_IMAGE:-vllm/vllm-router:nightly-20260617-e667ebb}"
+VLLM_ROUTER_IMAGE="${VLLM_ROUTER_IMAGE:-vllm/vllm-router:nightly-20260629-e667ebb}"
 ROUTER_CONT_NAME="router_vllm_${SANITIZED_USER}_${SLURM_JOB_ID}"
 export RUN_FILE_FULL="$WS_PATH/${RUN_FILE}"
 
@@ -532,6 +532,7 @@ if [[ \"$ENGINE\" == \"vllm-disagg\" && \"$ROUTER_TYPE\" == \"vllm-router\" && \
     \$DOCKER_CMD run -d \
         --name \"$ROUTER_CONT_NAME\" \
         --network host \
+        --ulimit nofile=1048576:1048576 \
         -v /tmp:/run_logs \
         \"$VLLM_ROUTER_IMAGE\" \
         bash -lc \"mkdir -p /run_logs/slurm_job-${SLURM_JOB_ID} && exec vllm-router \

configs/amd-master.yaml

@@ -1177,7 +1177,7 @@ dsr1-fp8-mi355x-sglang-disagg-mtp:
           - "DECODE_MTP_SIZE=2"
 
 kimik2.5-fp4-mi355x-vllm-disagg:
-  image: vllm/vllm-openai-rocm:nightly-bf610c2f56764e1b30bc6065f4ceace3d6e59036
+  image: vllm/vllm-openai-rocm:v0.24.0
   model: amd/Kimi-K2.5-MXFP4
   model-prefix: kimik2.5
   runner: mi355x-disagg
@@ -1200,7 +1200,6 @@ kimik2.5-fp4-mi355x-vllm-disagg:
           dp-attn: false
           additional-settings:
           - "PREFILL_NODES=1"
-          - "VLLM_MORIIO_CONNECTOR_READ_MODE=1"
         decode:
           num-worker: 2
           tp: 8
@@ -1221,7 +1220,6 @@ kimik2.5-fp4-mi355x-vllm-disagg:
           dp-attn: false
           additional-settings:
           - "PREFILL_NODES=1"
-          - "VLLM_MORIIO_CONNECTOR_READ_MODE=1"
         decode:
           num-worker: 2
           tp: 8

perf-changelog.yaml

@@ -4621,6 +4621,13 @@
     - "1k1k: 1 wide-EP max-throughput + 6 low-latency (DEP16/DEP32 + TP=4) topologies."
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1780
 
+- config-keys:
+    - kimik2.5-fp4-mi355x-vllm-disagg
+  description:
+    - "Bump image to vllm/vllm-openai-rocm:v0.24.0 (MoRI-IO fixes upstreamed in vllm#40344)"
+    - "Rename a2a backend mori to mori_low_latency; set read_mode via kv_connector_extra_config instead of VLLM_MORIIO_CONNECTOR_READ_MODE env var"
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1585
+
 - config-keys:
     - dsr1-fp4-b200-dynamo-sglang-mtp
   description: