Merge branch 'main' into saemal/py_mscclpp

microsoft · Nov 8, 2023 · d84a9f1 · d84a9f1
2 parents 23cbb94 + 0863e86
commit d84a9f1
Show file tree

Hide file tree

Showing 5 changed files with 22 additions and 22 deletions.
diff --git a/.azure-pipelines/multi-nodes-test.yml b/.azure-pipelines/multi-nodes-test.yml
@@ -28,7 +28,7 @@ jobs:
         curl -L https://github.com/Kitware/CMake/releases/download/v3.26.4/cmake-3.26.4-linux-x86_64.tar.gz -o /tmp/cmake-3.26.4-linux-x86_64.tar.gz
         tar xzf /tmp/cmake-3.26.4-linux-x86_64.tar.gz -C /tmp
         mkdir build && cd build
-        MPI_HOME=/usr/local/mpi /tmp/cmake-3.26.4-linux-x86_64/bin/cmake -DCMAKE_BUILD_TYPE=Release ..
+        MPI_HOME=/usr/local/mpi /tmp/cmake-3.26.4-linux-x86_64/bin/cmake -DCMAKE_BUILD_TYPE=Release -DBYPASS_PEERMEM_CHECK=ON ..
         make -j
         make pylib-copy
       workingDirectory: '$(System.DefaultWorkingDirectory)'
@@ -53,11 +53,11 @@ jobs:
     name: StartVMSS
     displayName: Start VMSS
     inputs:
-      azureSubscription: mscclpp
+      azureSubscription: msccl-it
       scriptType: bash
       scriptLocation: inlineScript
       inlineScript: |
-        az vmss start --name mscclpp-it-vmss --resource-group msccl-dev
+        az vmss start --name mscclit-vmss --resource-group msccl-IT
 
   - task: Bash@3
     name: DeployTestEnv
@@ -79,10 +79,10 @@ jobs:
         KeyFilePath=${SSHKEYFILE_SECUREFILEPATH}
         rm -rf output/*
         mkdir -p output
-        touch output/mscclpp-it-000000
-        tail -f output/mscclpp-it-000000 &
+        touch output/mscclit-000000
+        tail -f output/mscclit-000000 &
         CHILD_PID=$!
-        parallel-ssh -t 0 -H mscclpp-it-000000 -l azureuser -x "-i ${KeyFilePath}" \
+        parallel-ssh -t 0 -H mscclit-000000 -l azureuser -x "-i ${KeyFilePath}" \
         -O $SSH_OPTION -o output 'sudo docker exec -t mscclpp-test bash /root/mscclpp/run_tests.sh mscclpp-test'
         kill $CHILD_PID
 
@@ -98,10 +98,10 @@ jobs:
         KeyFilePath=${SSHKEYFILE_SECUREFILEPATH}
         rm -rf output/*
         mkdir -p output
-        touch output/mscclpp-it-000000
-        tail -f output/mscclpp-it-000000 &
+        touch output/mscclit-000000
+        tail -f output/mscclit-000000 &
         CHILD_PID=$!
-        parallel-ssh -t 0 -H mscclpp-it-000000 -l azureuser -x "-i ${KeyFilePath}" \
+        parallel-ssh -t 0 -H mscclit-000000 -l azureuser -x "-i ${KeyFilePath}" \
         -O $SSH_OPTION -o output 'sudo docker exec -t mscclpp-test bash /root/mscclpp/run_tests.sh mp-ut'
         kill $CHILD_PID
 
@@ -117,10 +117,10 @@ jobs:
         KeyFilePath=${SSHKEYFILE_SECUREFILEPATH}
         rm -rf output/*
         mkdir -p output
-        touch output/mscclpp-it-000000
-        tail -f output/mscclpp-it-000000 &
+        touch output/mscclit-000000
+        tail -f output/mscclit-000000 &
         CHILD_PID=$!
-        parallel-ssh -t 0 -H mscclpp-it-000000 -l azureuser -x "-i ${KeyFilePath}" \
+        parallel-ssh -t 0 -H mscclit-000000 -l azureuser -x "-i ${KeyFilePath}" \
         -O $SSH_OPTION -o output 'sudo docker exec -t mscclpp-test bash /root/mscclpp/run_tests.sh pytests'
         kill $CHILD_PID
 
@@ -129,8 +129,8 @@ jobs:
     displayName: Deallocate VMSS
     condition: always()
     inputs:
-      azureSubscription: mscclpp
+      azureSubscription: msccl-it
       scriptType: bash
       scriptLocation: inlineScript
       inlineScript: |
-        az vmss deallocate  --name mscclpp-it-vmss --resource-group msccl-dev
+        az vmss deallocate  --name mscclit-vmss --resource-group msccl-IT
diff --git a/test/deploy/config b/test/deploy/config
@@ -1,8 +1,8 @@
-Host mscclpp-it-000000
+Host mscclit-000000
   Port 22345
   IdentityFile /root/mscclpp/sshkey
   StrictHostKeyChecking no
-Host mscclpp-it-000001
+Host mscclit-000001
   Port 22345
   IdentityFile /root/mscclpp/sshkey
   StrictHostKeyChecking no
diff --git a/test/deploy/hostfile b/test/deploy/hostfile
@@ -1,2 +1,2 @@
-azureuser@mscclpp-it-000000
-azureuser@mscclpp-it-000001
+azureuser@mscclit-000000
+azureuser@mscclit-000001
diff --git a/test/deploy/hostfile_mpi b/test/deploy/hostfile_mpi
@@ -1,2 +1,2 @@
-mscclpp-it-000000
-mscclpp-it-000001
+mscclit-000000
+mscclit-000001
diff --git a/test/deploy/run_tests.sh b/test/deploy/run_tests.sh
@@ -52,12 +52,12 @@ function run_mp_ut()
   echo "============Run multi-process unit tests on 2 nodes (np=2, npernode=1)========================="
   /usr/local/mpi/bin/mpirun -allow-run-as-root -tag-output -np 2 --bind-to numa \
   -hostfile /root/mscclpp/hostfile_mpi -x MSCCLPP_DEBUG=WARN -x LD_LIBRARY_PATH=/root/mscclpp/build:$LD_LIBRARY_PATH \
-  -npernode 1 /root/mscclpp/build/test/mp_unit_tests -ip_port mscclpp-it-000000:20003
+  -npernode 1 /root/mscclpp/build/test/mp_unit_tests -ip_port mscclit-000000:20003
 
   echo "============Run multi-process unit tests on 2 nodes (np=16, npernode=8)========================="
   /usr/local/mpi/bin/mpirun -allow-run-as-root -tag-output -np 16 --bind-to numa \
   -hostfile /root/mscclpp/hostfile_mpi -x MSCCLPP_DEBUG=WARN -x LD_LIBRARY_PATH=/root/mscclpp/build:$LD_LIBRARY_PATH \
-  -npernode 8 /root/mscclpp/build/test/mp_unit_tests -ip_port mscclpp-it-000000:20003
+  -npernode 8 /root/mscclpp/build/test/mp_unit_tests -ip_port mscclit-000000:20003
 }
 
 function run_pytests()