ci: run the failure-path load test nightly instead of on every pull request

A load test that measures RSS and latency over a dozen requests belongs on a schedule, not in
the per-PR matrix. Drop the unit-test shard and add a nightly workflow (09:00 UTC, plus manual
dispatch) that runs it alone against the default branch. It needs no secrets, database or
network beyond localhost.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014ZDULyJPp17ZFiJenRxs2T

.github/workflows/nightly_failure_path_load.yml

@@ -0,0 +1,50 @@
+name: "Nightly Failure-Path Load Test"
+
+on:
+  schedule:
+    - cron: "0 9 * * *"
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  failure-path-load:
+    if: github.event_name != 'schedule' || github.repository == 'BerriAI/litellm'
+    runs-on: ubuntu-latest
+    timeout-minutes: 20
+    steps:
+      - uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # v4.3.0
+        with:
+          persist-credentials: false
+
+      - name: Set up Python
+        uses: actions/setup-python@a26af69be951a213d495a4c3e4e4022e16d87065 # v5.6.0
+        with:
+          python-version: "3.12"
+
+      - name: Set up uv
+        uses: ./.github/actions/setup-uv-with-retries
+        with:
+          version: "0.10.9"
+
+      - name: Cache the Rust build
+        uses: ./.github/actions/cache-cargo-build
+
+      - name: Install dependencies
+        run: |
+          .github/scripts/uv_sync_with_retries.sh --frozen --group ci --group proxy-dev --extra proxy
+
+      - name: Cache Prisma binaries
+        uses: ./.github/actions/cache-prisma-binaries
+
+      - name: Generate Prisma client
+        run: |
+          uv run --no-sync prisma generate --schema litellm/proxy/schema.prisma
+
+      - name: Run the failure-path load test
+        env:
+          LITELLM_LOG: WARNING
+        run: |
+          uv run --no-sync pytest tests/load_tests/test_failing_dependencies_load.py -v --tb=short -rA \
+            -o log_cli=true -o log_cli_level=INFO

.github/workflows/test-unit.yml

@@ -244,14 +244,6 @@ jobs:
             reruns: 2
             timeout-minutes: 20
             job-timeout-minutes: 60
-
-          - shard: failure-path-load
-            artifact-name: failure-path-load
-            test-path: "tests/load_tests/test_failing_dependencies_load.py"
-            workers: 1
-            reruns: 1
-            timeout-minutes: 10
-            job-timeout-minutes: 20
     uses: ./.github/workflows/_test-unit-base.yml
     with:
       test-path: ${{ matrix.test-path }}