Menu
Benchmark and autoscaling test harness for Huawei Cloud CSS (OpenSearch). Use when the task is to evaluate CSS performance, run load tests, validate query quality, test data-node horizontal autoscaling, or generate consolidated benchmark reports for CSS clusters.
| name | css-autoscaling-benchmark-skill |
| description | Benchmark and autoscaling test harness for Huawei Cloud CSS (OpenSearch). Use when the task is to evaluate CSS performance, run load tests, validate query quality, test data-node horizontal autoscaling, or generate consolidated benchmark reports for CSS clusters. |
Comprehensive benchmark and autoscaling test harness for Huawei Cloud CSS (Cloud Search Service / OpenSearch), built around a ColombiaEats food delivery use case with vector similarity search, geolocation, and cross-query patterns.
run_complete_benchmark.sh <-- Main entry point (7-phase pipeline)
├── config.py <-- Centralized .env configuration
├── setup_css_index.py <-- Phase 1: Index creation
├── ingesta_benchmark.py <-- Phase 2: Bulk ingestion
├── evaluaciones_benchmark.py <-- Phase 3: Query quality evaluation
├── cross_query_benchmark.py <-- Phase 4: CSS + RDS cross-queries
├── run_performance_tests.sh <-- Phase 5: Locust load tests
│ └── locustfile.py
├── monitor_autoscaling.py <-- Phase 6: Cluster monitoring
├── phase5_controller.py <-- Phase 6: Aggressive load to trigger scaling
├── data_node_autoscaler.py <-- Autoscaler orchestrator
├── css_monitor.py <-- Data node metrics collector
├── scaling_engine.py <-- Hysteresis-based scaling decisions
├── huawei_css_api.py <-- Huawei Cloud CSS SDK wrapper
├── generate_report.py <-- Phase 7: Consolidated report
├── apis_test.py <-- API endpoint validation
├── load_generator.py <-- Standalone load generator
└── stress_test_wrapper.py <-- Stress test orchestration
config.py to control concurrency and ramp rate.| Phase | Script | Description |
|---|---|---|
| 1 | setup_css_index.py | Create indices with mappings for vectors, geolocation, and text |
| 2 | ingesta_benchmark.py | Bulk ingest ColombiaEats documents with embeddings |
| 3 | evaluaciones_benchmark.py | Evaluate query quality (recall, precision, latency) |
| 4 | cross_query_benchmark.py | Cross-query CSS + simulated RDS for pricing/availability |
| 5 | run_performance_tests.sh | Locust load tests with configurable concurrency |
| 6 | monitor_autoscaling.py | Monitor cluster; trigger scaling with aggressive load |
| 7 | generate_report.py | Consolidated report with all metrics and scaling events |
# Configure
cp .env.example .env
# Edit .env with CSS endpoint, AK/SK, region
# Install dependencies
pip install -r requirements.txt
# Run full 7-phase benchmark
bash run_complete_benchmark.sh
# Or run individual phases
python3 setup_css_index.py
python3 ingesta_benchmark.py
python3 evaluaciones_benchmark.py
| Symptom | Cause | Fix |
|---|---|---|
CSS.0065 disk sold out | COMMON disk unavailable | Change css_volume_type to HIGH in config |
| Locust connection timeout | CSS endpoint unreachable or SG blocked | Verify endpoint and security group rules |
| Autoscaling not triggered | Load too low or cooldown period active | Increase phase5_controller.py concurrency; check cluster cooldown |
| Ingestion rate drops | Bulk request too large | Reduce chunk_size in config.py |
Use when analyzing or migrating Databricks CDC/Delta workflow scripts into a Huawei Cloud Chile MRS + OBS + Apache Hudi demo, including synthetic CDC data generation, notebook-triggered automation, Delta-to-Hudi replacement, deployment troubleshooting, and smoke validation. Also use when continuing the dockone ExampleApp demo package under the Codex outputs folder.
Configure Claude Code to use Huawei Cloud MaaS or ModelArts MaaS through an OpenAI-compatible endpoint, and optionally add Z.ai web-search-prime MCP search. Use when Codex needs to add a side-by-side claude-glm command that routes to Huawei MaaS glm-5.1 while preserving the original claude command on Anthropic, migrate claude itself to Huawei MaaS, install or configure claude-code-router, set API_KEY-based authentication, adjust context length, verify that Claude Code is actually backed by MaaS, or configure Z.ai MCP search with Z_API_KEY.
Deploy, configure, validate, troubleshoot, or extend an OpenAI-compatible API proxy backed by PostgreSQL, Prometheus, and Grafana, routing through Huawei ModelArts MaaS (ap-southeast-1) with multi-key load balancing. TRIGGER when the task involves LiteLLM proxy deployment, Docker Compose stack with litellm_config.yaml, Huawei MaaS model routing, virtual key or budget management, Prometheus/Grafana observability for LLM traffic, custom_callbacks.py TTFT/TPOT/ITL metrics, multi-key load balancing, or any reference to `LITELLM_MASTER_KEY`, `HUAWEI_MAAS_API_KEY`, or `docker compose` with this stack.
Bootstrap AI coding stack: deploy LiteLLM proxy (via LiteLLM-Huawei-MaaS-Proxy skill), install opencode + oh-my-opencode-slim, mint virtual key, wire everything. Supports multi-key MaaS load balancing. TRIGGER on: opencode + Huawei MaaS setup, full-stack bootstrap, oh-my-opencode-slim-huawei-maas, deploy-litellm.
End-to-end screenshot-to-dashboard pipeline powered by MaaS vision models. Analyze a dashboard screenshot via OpenRouter free vision models, extract design tokens with anydesign, generate design.md + design-tokens.json, plan via CCPM, implement React+Ant Design+ECharts dashboard code, and iterate with screenshot comparison until ≥85% similarity. TRIGGER when the user provides a dashboard screenshot and wants to reconstruct it as a working web app, or says 'rebuild this dashboard', 'convert this screenshot to code', 'reproduce this BI dashboard', 'screenshot to React dashboard'. Do NOT use for: non-dashboard screenshots, Figma-to-code without screenshot, general UI development without a reference image.
Disciplined diagnosis loop for hard bugs and performance regressions. Reproduce → minimise → hypothesise → instrument → fix → regression-test. Use when user says "diagnose this" / "debug this", reports a bug, says something is broken/throwing/failing, or describes a performance regression.