【运维发布】事故定级：P0 到 P3 的响应标准

---事故的定级：影响面与时长的矩阵
function Severity.classify(incident)
    local scope = incident.affectedPlayers   -- 受影响玩家数
    local duration = incident.durationMin    -- 预计持续
    -- P0：全服不可用或资损
    if scope >= 0.8 or incident.fundLoss then
        return 'P0'
    end
    -- P1：核心功能受损
    if incident.coreBroken or scope >= 0.3 then
        return 'P1'
    end
    -- P2：非核心受损，有替代方案
    if scope >= 0.1 then
        return 'P2'
    end
    return 'P3'
end

---响应的流程：定级后的标准动作
local RESPONSE = {
    P0 = { ack = 5,  update = 15, team = '全员' },
    P1 = { ack = 15, update = 30, team = '值班+模块负责' },
    P2 = { ack = 60, update = 120, team = '模块负责' },
    P3 = { ack = 240, update = 480, team = '模块负责' },
}

function Severity.respond(level, incident)
    local std = RESPONSE[level]
    -- 确认时限内的首次响应
    if os.time() - incident.at > std.ack * 60 then
        Ops.alert('响应超时', level, incident.id)
    end
    -- 周期性的进展通报
    incident.nextUpdate = os.time() + std.update * 60
end