[TRTLLM-7141][infra] Use repo mirrors to avoid intermittent network failures (NVIDIA#6836)

Signed-off-by: Yanchao Lu <[email protected]>
Co-authored-by: Zhanrui Sun <[email protected]>
Signed-off-by: Wangshanshan <[email protected]>

Author: 2 people

cpp/CMakeLists.txt

@@ -495,6 +495,17 @@ if(ENABLE_UCX)
   if(NOT ${ucx_FOUND})
     set(ENABLE_UCX 0)
   else()
+    if(DEFINED ENV{GITHUB_MIRROR} AND NOT "$ENV{GITHUB_MIRROR}" STREQUAL "")
+      if(EXISTS "${3RDPARTY_DIR}/ucxx/fetch_rapids.cmake")
+        file(READ "${3RDPARTY_DIR}/ucxx/fetch_rapids.cmake" FILE_CONTENTS)
+        string(
+          REPLACE "https://raw.githubusercontent.com/rapidsai/rapids-cmake"
+                  "$ENV{GITHUB_MIRROR}/rapidsai/rapids-cmake/raw/refs/heads"
+                  FILE_CONTENTS "${FILE_CONTENTS}")
+        file(WRITE "${3RDPARTY_DIR}/ucxx/fetch_rapids.cmake" "${FILE_CONTENTS}")
+        message(WARNING "Replace UCXX fetch_rapids.cmake with internal mirror")
+      endif()
+    endif()
     # installing ucxx via add_subdirectory results in strange cudart linking
     # error, thus using their installation script to isolate the installation
     # process until the issue is understood. And always trigger the build so

jenkins/L0_Test.groovy

@@ -2075,6 +2075,11 @@ def launchTestJobs(pipeline, testFilter, dockerNode=null)
                 checkPipStage = true
             }
 
+            if (cpu_arch == AARCH64_TRIPLE && values[5] != DLFW_IMAGE) {
+                checkPipStage = false
+                echo "Skip pip install sanity check due to https://nvbugs/5453827"
+            }
+
             if (checkPipStage) {
                 stage("Run LLMAPI tests") {
                     pipInstallSanitySpec = createKubernetesPodConfig(values[5], gpu_type, k8s_arch)