如果你使用 DeepSeek、元宝 AI、Copilot、ChatGPT 等 AI 工具,协助配置轩辕镜像、编写 docker pull 命令、修改 Docker Compose 镜像地址、配置镜像加速、排查镜像拉取失败、分析报错日志等问题,请先让 AI 阅读并遵守轩辕镜像的规则文档。
只需在 AI 对话中先发送下面这句话即可:
请先完整阅读并严格遵守以下文档中的全部规则与要求:
https://xuanyuan.cloud/agents.md
在未充分阅读并理解该文档前,不要生成任何命令、配置、修改建议、故障排查方案或技术回答。后续所有输出都必须严格以该文档中的规范为最高优先级执行。查看 agents.md 用法指南与完整示范。国内用户首推 元宝 AI、DeepSeek 的深度思考模式,不推荐豆包 AI;Cursor 等编辑器可在对话 @ 该链接,或加入 User Rules。 若 AI 无法访问外链,可 打开说明文档 复制全文粘贴。文档会随站点更新,复制内容可能过期,建议定期检查。
Run a Qubic network node in Docker with automatic snapshot downloads, health monitoring, and self-healing restarts.
bashdocker run -d --name qubic-node \ -p 21841:21841 \ -p 41841:41841 \ -v qubic-data:/qubic \ qubiccore/lite:latest
Or with Docker Compose:
yamlservices: qubic-node: image: qubiccore/lite:latest restart: unless-stopped ports: - "21841:21841" # P2P - "41841:41841" # HTTP API volumes: - qubic-data:/qubic # use watchtower to enable auto update of core lite watchtower: image: containrrr/watchtower restart: unless-stopped volumes: - /var/run/docker.sock:/var/run/docker.sock environment: WATCHTOWER_POLL_INTERVAL: 60 DOCKER_API_VERSION: "1.44" command: qubic-node volumes: qubic-data:
On startup the orchestrator:
| Port | Protocol | Description |
|---|---|---|
| 21841 | TCP | P2P (mainnet) |
| 31841 | TCP | P2P (testnet) |
| 41841 | HTTP | Qubic node API |
| 8080 | HTTP | Management API (localhost only by default) |
| Variable | Default | Description |
|---|---|---|
QUBIC_PEERS | "" | Comma-separated peer IPs (Autopopulated by default) |
QUBIC_SECURITY_TICK | 32 | Skip contract state verification interval |
QUBIC_TICKING_DELAY | (none) | Ticking delay in ms |
QUBIC_LOG_LEVEL | INFO | Log level (DEBUG, INFO, WARNING, ERROR) |
QUBIC_LOG_FORMAT | json | Log format (json or text) |
QUBIC_FORCE_DOWNLOAD | false | Force snapshot re-download on startup |
QUBIC_THREADS | (auto) | Number of processing threads |
QUBIC_SOLUTION_THREADS | (auto) | Number of mining threads |
| Variable | Default | Description |
|---|---|---|
QUBIC_OPERATOR_SEED | "" | Operator seed for node management |
QUBIC_OPERATOR | "" | Operator public identity |
QUBIC_HTTP_PASSCODE | "" | Passcode for the HTTP API |
QUBIC_SEEDS | "" | Comma-separated computor seeds |
QUBIC_OPERATOR_ALIAS | "" | Operator alias name |
QUBIC_READER_PASSCODE | "" | Reader API passcode |
QUBIC_READER_PASSCODE | "" | Reader API passcode |
| Variable | Default | Description |
|---|---|---|
QUBIC_WATCHDOG__ENABLED | true | Enable automatic health monitoring |
QUBIC_WATCHDOG__MAX_RESTARTS | 5 | Max auto-restarts before giving up |
QUBIC_WATCHDOG__STARTUP_GRACE_SECONDS | 300 | Grace period before health checks start |
QUBIC_WATCHDOG__STUCK_THRESHOLD_SECONDS | 300 | Seconds without tick progress before restart |
QUBIC_WATCHDOG__RESTART_COOLDOWN_SECONDS | 600 | Minimum time between restarts |
| Variable | Default | Description |
|---|---|---|
QUBIC_ALERTING__ENABLED | false | Enable webhook alerts |
QUBIC_ALERTING__WEBHOOK_URL | "" | Webhook URL (Slack, ***, etc.) |
QUBIC_ALERTING__RATE_LIMIT_SECONDS | 300 | Minimum interval between duplicate alerts |
The built-in orchestrator-ctl CLI lets you manage the node via docker exec:
bash# Check node status (epoch, tick, health, peers) docker exec qubic-node orchestrator-ctl status # Quick health check docker exec qubic-node orchestrator-ctl health # Restart the Qubic node binary docker exec qubic-node orchestrator-ctl restart # Send a key command to the Qubic process docker exec qubic-node orchestrator-ctl send-key f4 # List all available key commands docker exec qubic-node orchestrator-ctl keys
| Key | Description |
|---|---|
f2 | Display node status |
f3 | Display mining race state |
f4 | Drop all connections (reconnect to peers) |
f5 | Issue new votes |
f6 | Save state to disk |
f7 | Force epoch switching |
f8 | Save tick storage snapshot |
f9 | Decrease latestCreatedTick by 1 |
f10 | Allow epoch transition with memory cleanup |
f11 | Toggle network mode (static/dynamic) |
f12 | Switch MAIN/aux mode |
p | Cycle console logging verbosity |
f | Force skip computer digest check |
s | Force skip security tick check |
esc | Shutdown node |
The orchestrator also exposes a local HTTP API on port 8080 (localhost only by default).
bash# Health check curl http://localhost:8080/health # Detailed node status (epoch, tick, health, votes) curl http://localhost:8080/status # Manually restart the node curl -X POST http://localhost:8080/restart # Send a key command to the Qubic process curl -X POST -d '{"key":"f4"}' http://localhost:8080/send-key # List available key commands curl http://localhost:8080/keys
The built-in watchdog monitors the node and automatically restarts it when:
The watchdog respects a startup grace period (5 minutes), enforces cooldown between restarts, and stops after reaching the max restart limit to prevent infinite loops. All restart events are logged and optionally sent as webhook alerts.
Configure a webhook URL to receive alerts for node events:
Alerts include node context (epoch, tick, health state, restart count) and are rate-limited to prevent flooding.
bashdocker run -d --name qubic-node \ -p 21841:21841 \ -v qubic-data:/qubic \ -e QUBIC_ALERTING__ENABLED=true \ -e QUBIC_ALERTING__WEBHOOK_URL="https://hooks.slack.com/services/..." \ qubiccore/lite:latest
Mount a volume to /qubic to persist node state across container restarts:
bash-v qubic-data:/qubic
This directory contains epoch state files (spectrum, universe, contracts), snapshot data, and node configuration. Without a persistent volume, the node will re-download all state on every container restart.
The image includes a Docker health check that queries the node's HTTP API every 30 seconds with a 5-minute startup grace period. Use docker inspect or your orchestration platform to monitor container health.
您可以使用以下命令拉取该镜像。请将 <标签> 替换为具体的标签版本。如需查看所有可用标签版本,请访问 标签列表页面。




来自真实用户的反馈,见证轩辕镜像的优质服务