备份与灾难恢复设置

部署 4 个代理：scheduler（调度器）、primary backup（主备份）、offsite secondary（异地副本）和 restore tester（恢复测试器）。

角色

-backup-sched

-backup-primary

-backup-secondary

-restore-tester

设置流程

步骤 1： 向用户询问所选角色及前缀（prefix）。

步骤 2： 安装对应技能：

# scheduler：
clawhub install pilot-cron pilot-task-chain pilot-audit-log pilot-slack-bridge
# primary：
clawhub install pilot-backup pilot-archive pilot-compress pilot-share
# secondary：
clawhub install pilot-sync pilot-verify pilot-health pilot-share
# tester：
clawhub install pilot-backup pilot-verify pilot-health pilot-alert

步骤 3： 设置主机名，并将配置清单写入 ~/.pilot/setups/backup-disaster-recovery.json。

步骤 4： 沿链路完成握手：scheduler ↔ primary ↔ secondary ↔ tester ↔ scheduler。

各角色的清单模板

scheduler

{
  "setup": "backup-disaster-recovery", "role": "scheduler", "role_name": "Backup Scheduler",
  "hostname": "
-backup-sched",
  "skills": {
    "pilot-cron": "按周期（每小时、每日、每周）调度备份任务。",
    "pilot-task-chain": "编排备份 → 同步 → 校验 → 报告全流程。",
    "pilot-audit-log": "记录所有备份生命周期事件。",
    "pilot-slack-bridge": "将备份报告及失败告警发送至 Slack。"
  },
  "data_flows": [
    { "direction": "send", "peer": "
-backup-primary", "port": 1002, "topic": "backup-start", "description": "触发备份任务" },
    { "direction": "receive", "peer": "
-restore-tester", "port": 1002, "topic": "restore-result", "description": "恢复测试结果" }
  ],
  "handshakes_needed": ["
-backup-primary", "
-backup-secondary", "
-restore-tester"]
}

primary
{
  "setup": "backup-disaster-recovery", "role": "primary", "role_name": "Primary Backup",
  "hostname": "
-backup-primary",
  "skills": {
    "pilot-backup": "创建数据库及文件系统备份。",
    "pilot-archive": "归档备份（含元数据与保留策略）。",
    "pilot-compress": "在传输前压缩归档包。",
    "pilot-share": "将压缩后的归档包推送至 secondary。"
  },
  "data_flows": [
    { "direction": "receive", "peer": "
-backup-sched", "port": 1002, "topic": "backup-start", "description": "接收备份触发指令" },
    { "direction": "send", "peer": "
-backup-secondary", "port": 1001, "topic": "sync-backup", "description": "压缩后的归档包" }
  ],
  "handshakes_needed": ["
-backup-sched", "
-backup-secondary"]
}

secondary
{
  "setup": "backup-disaster-recovery", "role": "secondary", "role_name": "Offsite Replica",
  "hostname": "
-backup-secondary",
  "skills": {
    "pilot-sync": "从 primary 接收并同步备份归档包。",
    "pilot-verify": "校验接收到的备份归档包完整性校验和。",
    "pilot-health": "上报异地存储健康状态与容量信息。",
    "pilot-share": "按需向 restore tester 提供备份数据。"
  },
  "data_flows": [
    { "direction": "receive", "peer": "
-backup-primary", "port": 1001, "topic": "sync-backup", "description": "压缩后的归档包" },
    { "direction": "send", "peer": "
-restore-tester", "port": 1001, "topic": "restore-data", "description": "用于测试的备份数据" },
    { "direction": "send", "peer": "
-backup-sched", "port": 1002, "topic": "sync-confirmed", "description": "同步完成确认" }
  ],
  "handshakes_needed": ["
-backup-primary", "
-backup-sched", "
-restore-tester"]
}

tester
{
  "setup": "backup-disaster-recovery", "role": "tester", "role_name": "Restore Tester",
  "hostname": "
-restore-tester",
  "skills": {
    "pilot-backup": "在隔离环境中执行恢复测试。",
    "pilot-verify": "验证恢复后数据完整性（如行数、校验和）。",
    "pilot-health": "监控恢复测试环境健康状态。",
    "pilot-alert": "在恢复失败或数据损坏时发出告警。"
  },
  "data_flows": [
    { "direction": "receive", "peer": "
-backup-secondary", "port": 1001, "topic": "restore-data", "description": "待测试的备份数据" },
    { "direction": "send", "peer": "
-backup-sched", "port": 1002, "topic": "restore-result", "description": "测试结果" }
  ],
  "handshakes_needed": ["
-backup-secondary", "
-backup-sched"]
}

数据流向

scheduler → primary ：备份触发指令（端口 1002）
primary → secondary ：压缩后的归档包（端口 1001）
secondary → tester ：用于测试的备份数据（端口 1001）
tester → scheduler ：恢复测试结果（端口 1002）

工作流示例
# 在 scheduler 上执行：
pilotctl --json publish 
-backup-primary backup-start '{"job_id":"BK-315","type":"full","targets":["postgres","redis"]}'
# 在 primary 上执行：
pilotctl --json send-file 
-backup-secondary ./backups/BK-315-postgres.tar.gz
pilotctl --json publish 
-backup-secondary sync-backup '{"job_id":"BK-315","size_mb":1240}'
# 在 tester 上执行：
pilotctl --json publish 
-backup-sched restore-result '{"job_id":"BK-315","restore":"success","tables_verified":42}'

依赖项
需要 pilot-protocol 技能、pilotctl 可执行文件、clawhub 可执行文件，以及一个正在运行的守护进程（daemon）。