CI: Add AWS instance type to stats (#4687)

* CI: Add AWS instance type to stats

* Update scripts

* Collect stats

* Create dedicated GitHub action

* Revert "Create dedicated GitHub action"

This reverts commit cd28afa.

* Add fallback value

Author: oitel

.github/actions/collect-runner-stats/action.yml

@@ -17,6 +17,10 @@ inputs:
     required: false
     type: string
     default: ''
+  aws_instance_type:
+    required: false
+    type: string
+    default: ''
 
 outputs:
   job_id:
@@ -54,6 +58,7 @@ runs:
         CXX_COMPILER: ${{ inputs.cxx_compiler }}
         BUILD_CONFIG: ${{ inputs.build_config }}
         BUILD_SYSTEM: ${{ inputs.build_system }}
+        AWS_INSTANCE_TYPE: ${{ inputs.aws_instance_type }}
         STATS_FILE: "${{ runner.temp }}/RunnerSysStats-${{ steps.fetch_job_id.outputs.job_id }}.json"
       run: ""
 

.github/workflows/build-test-emscripten.yml

@@ -51,6 +51,25 @@ jobs:
             mkdir ${GITHUB_WORKSPACE}
           fi
 
+        # https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/instancedata-data-retrieval.html
+      - name: Get AWS instance type
+        run: |
+          TOKEN=$(curl -fsS --max-time 2 \
+            -X PUT "http://169.254.169.254/latest/api/token" \
+            -H "X-aws-ec2-metadata-token-ttl-seconds: 10" || true)
+
+          AWS_INSTANCE_TYPE=$(curl -fsS --max-time 2 \
+            -H "X-aws-ec2-metadata-token: $TOKEN" \
+            "http://169.254.169.254/latest/meta-data/instance-type" || true)
+          
+          # if not an EC2 instance
+          if [[ -z "$AWS_INSTANCE_TYPE" ]]; then
+            AWS_INSTANCE_TYPE="self-hosted"
+          fi
+
+          echo "AWS_INSTANCE_TYPE=$AWS_INSTANCE_TYPE" >> $GITHUB_ENV
+          echo $AWS_INSTANCE_TYPE
+
       - name: Checkout
         uses: actions/checkout@v4
 
@@ -63,6 +82,7 @@ jobs:
           target_arch: wasm
           cxx_compiler: emcc
           build_config: Release
+          aws_instance_type: ${{ env.AWS_INSTANCE_TYPE }}
 
       - name: Checkout third-party submodules
         run: |

.github/workflows/build-test-ubuntu-arm64.yml

@@ -69,6 +69,25 @@ jobs:
             mv ${GITHUB_WORKSPACE} ${GITHUB_WORKSPACE}_${RANDOM}
             mkdir ${GITHUB_WORKSPACE}
           fi
+          
+        # https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/instancedata-data-retrieval.html
+      - name: Get AWS instance type
+        run: |
+          TOKEN=$(curl -fsS --max-time 2 \
+            -X PUT "http://169.254.169.254/latest/api/token" \
+            -H "X-aws-ec2-metadata-token-ttl-seconds: 10" || true)
+
+          AWS_INSTANCE_TYPE=$(curl -fsS --max-time 2 \
+            -H "X-aws-ec2-metadata-token: $TOKEN" \
+            "http://169.254.169.254/latest/meta-data/instance-type" || true)
+          
+          # if not an EC2 instance
+          if [[ -z "$AWS_INSTANCE_TYPE" ]]; then
+            AWS_INSTANCE_TYPE="self-hosted"
+          fi
+
+          echo "AWS_INSTANCE_TYPE=$AWS_INSTANCE_TYPE" >> $GITHUB_ENV
+          echo $AWS_INSTANCE_TYPE
 
       - name: Checkout
         uses: actions/checkout@v4
@@ -82,6 +101,7 @@ jobs:
           target_arch: arm64
           cxx_compiler: ${{ matrix.cxx-compiler }}
           build_config: ${{ matrix.config }}
+          aws_instance_type: ${{ env.AWS_INSTANCE_TYPE }}
 
       - name: Checkout third-party submodules
         run: |

scripts/devops/collect_ci_stats.py

@@ -56,6 +56,7 @@ def parse_job(job: dict):
             'runner_cpu_count':  runner_stats['cpu_count'],
             'runner_ram_mb':     runner_stats['ram_mb'],
             'build_system':      runner_stats['build_system'],
+            'aws_instance_type': runner_stats['aws_instance_type'],
         }
     except:
         print("Something went wrong while parsing the job/runner info. Debug info:")

scripts/devops/collect_runner_sys_stats.py

@@ -62,6 +62,8 @@ def get_compiler_id(compiler_path):
         if not build_system:
             build_system = "msbuild" if compiler_id.startswith("msvc") else "cmake"
 
+        aws_instance_type = os.environ.get('AWS_INSTANCE_TYPE', '').lower()
+
         results = {
             'target_os': os.environ.get('TARGET_OS'),
             'target_arch': os.environ.get('TARGET_ARCH'),
@@ -70,6 +72,7 @@ def get_compiler_id(compiler_path):
             'cpu_count': cpu_count,
             'ram_mb': ram_amount,
             'build_system': build_system,
+            'aws_instance_type': aws_instance_type or None,
         }
         with open(os.environ['STATS_FILE'], 'w') as f:
             json.dump(results, f)