Postgres - Listen/Notify构建轻量级发布订阅系统

news2026/4/10 14:36:31
概述原先设计一个内部系统的消息模块和缓存模块时只有一个Postgres依赖。想着没多大用户量没必要额外安装Redis徒增运维工夫。缓存好解决配个UNLOGGED表即可。吭吭哧哧琢磨怎么用数据表实现消息的时候发现PostgreSQL 提供了内置命令LISTEN和NOTIFY用于在数据库服务器和连接的客户端之间实现异步通信。这个 PostgreSQL 特有的扩展功能使得数据库可以作为一个轻量级的消息队列MQ系统使用允许应用程序从数据库中生成事件并由其他客户端实时响应。于是一拍即合上手体验一下。核心特性轻量级实现无需额外的消息中间件直接利用 PostgreSQL 内置功能异步通信支持发布-订阅模式实现解耦的组件通信内存高效通道Channel是纯内存对象不占用磁盘空间零配置无需预先创建或管理通道随用随建适用场景实时仪表盘数据变更时实时推送更新缓存失效数据更新时通知缓存层刷新数据审计跟踪重要数据变更事件任务调度构建简单的分布式任务队列事件驱动架构实现微服务间的事件通信通道Channel机制重要特性Channel 是纯内存对象随LISTEN命令隐式创建当所有监听会话断开或执行UNLISTEN时自动回收无需手动创建或删除通道也不支持此操作消息传递模型PostgreSQL 的NOTIFY采用典型的无监听即丢弃机制没有监听者时消息不会入队不占用磁盘空间不消耗持久化内存消息仅在存在活跃监听者时传递基础使用psql 命令行示例-- 监听指定通道 LISTEN task_channel; -- 向通道发送消息 NOTIFY task_channel, 123456; -- 取消监听所有通道 UNLISTEN *; -- 查看当前监听的通道 SELECT pg_listening_channels(); -- 查看系统通知状态 SELECT * FROM pg_stat_activity WHERE backend_type client backend;动态消息生成标准的NOTIFY命令要求消息内容必须明确指定不支持动态字符串拼接。但可以使用pg_notify()函数来生成动态通知-- 使用 pg_notify 函数支持动态消息 SELECT pg_notify(my_channel, Hello, || World!); -- 带参数的动态消息 SELECT pg_notify(audit_channel, User || current_user || logged in at || now()::text);Python 实现示例项目结构├── main.py # 核心实现TaskWorker 和 TaskProducer ├── conf/ │ └── config.toml # 配置文件 └── pkg/ └── config/ # 配置管理模块安装依赖uv add psycopg[binary,pool]3.3.3配置管理首先通过配置文件管理数据库连接和通道设置# conf/config.toml [database.postgres] host 127.0.0.1 port 5432 user username password password dbname database_name pool_min_size 2 pool_max_size 10 channel task_channel # 默认通道名称配置模块的代码示例pkg/config/config.pyimport tomllib from typing import Any, Dict class BaseConfig: def __init__(self, cfg_file: str): self._cfg_file cfg_file self._data: Dict[str, Any] {} self._load_config() def _load_config(self) - None: if self._data: return try: with open(self._cfg_file, rb) as f: self._data tomllib.load(f) except FileNotFoundError: raise RuntimeError(f配置文件不存在: {self._cfg_file}) except Exception as e: raise RuntimeError(f加载配置文件失败: {e}) class PostgresConfigMixin(BaseConfig): def postgres_host(self) - str: return self._data.get(database, {}).get(postgres, {}).get(host, ) def postgres_port(self) - int: return self._data.get(database, {}).get(postgres, {}).get(port, 5432) def postgres_user(self) - str: return self._data.get(database, {}).get(postgres, {}).get(user, ) def postgres_password(self) - str: return self._data.get(database, {}).get(postgres, {}).get(password, ) def postgres_dbname(self) - str: return self._data.get(database, {}).get(postgres, {}).get(dbname, ) def postgres_pool_min_size(self) - int: return ( self._data.get(database, {}).get(postgres, {}).get(pool_min_size, 2) ) def postgres_pool_max_size(self) - int: return ( self._data.get(database, {}).get(postgres, {}).get(pool_max_size, 10) ) def postgres_channel(self) - str: return ( self._data.get(database, {}) .get(postgres, {}) .get(channel, default_channel) ) def get_postgres_dsn(self, hide_password: bool False) - str: 获取PostgreSQL连接DSN password self.postgres_password() if hide_password and password: password *** # psycopg3 使用标准的 PostgreSQL 连接字符串格式 return ( fpostgresql://{self.postgres_user()}:{password} f{self.postgres_host()}:{self.postgres_port()}/{self.postgres_dbname()} ) class LLMConfigMixin(BaseConfig): LLM配置Mixin def llm_model(self) - str: return self._data.get(llm, {}).get(model, ) def llm_base_url(self) - str: return self._data.get(llm, {}).get(base_url, ) def llm_api_key(self) - str: return self._data.get(llm, {}).get(api_key, ) class Config(PostgresConfigMixin, LLMConfigMixin): def __init__(self, cfg_file: str conf/config.toml): super().__init__(cfg_file) def reload(self) - None: 重新加载配置 self._data {} # 清空数据 self._load_config()核心组件实现1. 任务消费者TaskWorkerTaskWorker负责监听指定通道并处理接收到的任务# main.py - TaskWorker 类核心部分 import asyncio import json import signal from typing import Set from psycopg import AsyncConnection, Notify, sql from psycopg_pool import AsyncConnectionPool from pkg.config import cfg MAX_CONCURRENCY 10 class TaskWorker: def __init__(self, dsn: str, channel: str): self._dsn dsn self.channel channel self.pool: AsyncConnectionPool | None None self.listener_conn: AsyncConnection | None None self.sem asyncio.Semaphore(MAX_CONCURRENCY) self.active_tasks: Set[asyncio.Task] set() async def start(self) - None: self.pool AsyncConnectionPool( self._dsn, min_sizecfg.postgres_pool_min_size(), max_sizecfg.postgres_pool_max_size(), openFalse, # 延迟打开, 避免阻塞 ) await self.pool.open() # 独立监听连接防止 LISTEN 状态随连接回收丢失 self.listener_conn await AsyncConnection.connect(self._dsn, autocommitTrue) await self.listener_conn.execute( sql.SQL(LISTEN {}).format(sql.Identifier(self.channel)) ) print(fListening on channel: {self.channel}) try: async for notify in self.listener_conn.notifies(): if notify.channel self.channel and notify.payload: await self._dispatch_task(notify) except asyncio.CancelledError: print(Listener cancelled) except Exception as e: print(fListener error: {e}) finally: await self.stop() async def _dispatch_task(self, notify: Notify) - None: 接收通知并分发任务 task_info notify.payload.strip() try: task_data json.loads(task_info) # 假设 payload 是 JSON 格式的字符串 print(fReceived task notification: {task_data}) except json.JSONDecodeError: task_data {task_id: task_info} # 如果不是 JSON 格式使用原始字符串 print(fReceived non-JSON task notification: {task_data}) except Exception as e: print(fInvalid task ID received: {task_info}) return if not isinstance(task_data, dict) or task_id not in task_data: print( fMissing task_id in notification: {task_data}, or task_data is not a dict ) return task asyncio.create_task(self._process_task(task_data)) self.active_tasks.add(task) task.add_done_callback(self.active_tasks.discard) async def _process_task(self, task_data: dict) - None: 执行业务逻辑 if not self.pool: raise RuntimeError(Connection pool is not initialized) async with self.sem: async with self.pool.connection() as conn: try: await self._execute_business(task_data) print(f Task {task_data[task_id]} completed successfully) except Exception as e: print(fError processing task {task_data[task_id]}: {e}) # 这里可以添加重试逻辑或错误记录 await self._log_failure(conn, task_data[task_id], str(e)) async def _execute_business(self, task_data: dict) - None: 执行业务逻辑 print(f Processing task {task_data}...) await asyncio.sleep(5) # 模拟耗时操作 print(f Task {task_data} done.) async def _log_failure(self, conn: AsyncConnection, task_id: int, error_msg: str): 记录失败日志 try: # 记录失败日志到独立表便于后续重试或告警 print(fLogging failure for task {task_id}: {error_msg}) except Exception as e: print(fFailed to log error for task {task_id}: {e}) async def stop(self) - None: print(Stopping TaskWorker...) if self.active_tasks: await asyncio.gather(*self.active_tasks, return_exceptionsTrue) if self.listener_conn: await self.listener_conn.close() if self.pool: await self.pool.close() print(TaskWorker stopped gracefully.)2. 任务发布者TaskPublisherTaskPublisher负责向通道发布任务消息# main.py - TaskPublisher 类 class TaskPublisher: def __init__(self, dsn: str): self._dsn dsn self._pool: AsyncConnectionPool | None None async def start(self): if not self._pool: self._pool AsyncConnectionPool( self._dsn, min_sizecfg.postgres_pool_min_size(), max_sizecfg.postgres_pool_max_size(), openFalse, # 延迟打开, 避免阻塞 ) await self._pool.open() print(TaskPublisher started.) async def publish(self, channel: str, payload: dict): if not self._pool: raise RuntimeError(Connection pool is not initialized) async with self._pool.connection() as conn: try: payload_str json.dumps( payload, defaultstr ) # 将 dict 转换为 JSON 字符串 await conn.execute( sql.SQL(NOTIFY {}, {}).format( sql.Identifier(channel), sql.Literal(payload_str) ) ) print(f Published task to channel {channel}: {payload_str}) return True except Exception as e: print(fFailed to publish task to channel {channel}: {e}) return False async def publish_batch(self, channel: str, payloads: list[dict]): if not self._pool: raise RuntimeError(Connection pool is not initialized) count 0 async with self._pool.connection() as conn: for payload in payloads: try: payload_str json.dumps(payload, defaultstr) await conn.execute( sql.SQL(NOTIFY {}, {}).format( sql.Identifier(channel), sql.Literal(payload_str) ) ) print(f Published task to channel {channel}: {payload_str}) count 1 except Exception as e: print(fFailed to publish batch tasks to channel {channel}: {e}) return count async def stop(self): if self._pool: await self._pool.close() print(TaskPublisher stopped.) async def __aenter__(self): await self.start() return self async def __aexit__(self, exc_type, exc_val, exc_tb): await self.stop()调用演示以下是main.py中的演示部分# main.py async def run_worker(): worker TaskWorker(cfg.get_postgres_dsn(), cfg.postgres_channel()) task asyncio.create_task(worker.start()) return task async def run_publisher(): async with TaskPublisher(cfg.get_postgres_dsn()) as publisher: for i in range(1, 11): payload {task_id: i, data: fTask data {i}} await publisher.publish(cfg.postgres_channel(), payload) await asyncio.sleep(0.5) # 模拟发布间隔 async def main(): # worker TaskWorker(cfg.get_postgres_dsn(), cfg.postgres_channel()) # loop asyncio.get_running_loop() # stop_evt asyncio.Event() # for sig in (signal.SIGINT, signal.SIGTERM): # loop.add_signal_handler(sig, stop_evt.set) # listen_task asyncio.create_task(worker.start()) # await stop_evt.wait() # print(Shutdown signal received, stopping worker...) # listen_task.cancel() # await listen_task worker_task await run_worker() await asyncio.sleep(2) # 确保 worker 已经启动并监听 await run_publisher() # 发布任务 print(All tasks published successfully.) await asyncio.sleep(30) # 等待 worker 处理完任务 worker_task.cancel() # 停止 worker try: await worker_task except asyncio.CancelledError: pass if __name__ __main__: asyncio.run(main())使用场景扩展场景一实时数据同步async def sync_data_change(self, table_name: str, record_id: str, operation: str): 数据变更时发送同步通知 message f{table_name}:{record_id}:{operation} await self.publish(data_sync_channel, message)场景二分布式锁通知async def notify_lock_release(self, lock_name: str): 锁释放时通知等待者 await self.publish(distributed_lock_channel, fRELEASE:{lock_name})场景三缓存失效广播async def invalidate_cache(self, cache_key: str): 缓存失效时广播通知 await self.publish(cache_invalidation_channel, cache_key)注意事项技术限制消息大小NOTIFY 消息最大为 8000 字节无持久化消息不持久化重启后丢失无确认机制发送方无法知道消息是否被接收无顺序保证消息可能不按发送顺序到达生产环境建议监控告警实现通道监听状态监控错误处理添加完善的错误处理和日志记录备份机制重要消息应有备份存储性能测试在高负载下测试系统表现补充asyncpg版asyncpg是python连接postgres的纯异步驱动性能更好安装uv add asyncpg代码示例import asyncio import signal from typing import Set import asyncpg from pkg.config import cfg MAX_CONCURRENT_TASKS 5 # 最大并发任务数 class TaskWorker: def __init__(self, dsn: str, channel: str): self._dsn dsn self._channel channel self._pool: asyncpg.Pool | None None self._listener_conn: asyncpg.Connection | None None self._sem asyncio.Semaphore(MAX_CONCURRENT_TASKS) self._active_tasks: Set[asyncio.Task] set() async def start(self): 启动监听器 if not self._pool: self._pool await asyncpg.create_pool( dsnself._dsn, min_sizecfg.postgres_pool_min_size(), max_sizecfg.postgres_pool_max_size(), ) # 创建专用连接用于监听 if not self._listener_conn: self._listener_conn await asyncpg.connect(dsnself._dsn) await self._listener_conn.add_listener(self._channel, self._on_notify) print(fListening on channel: {self._channel}) try: # 保持监听状态 await asyncio.Future() except asyncio.CancelledError: pass finally: await self.stop() async def _on_notify( self, conn: asyncpg.Connection, pid: int, channel: str, payload: str ): 收到 NOTIFY 时的回调函数 if not payload: return task_id payload.strip() print(fReceived notification: {task_id} on channel: {channel}) # 提交到事件循环带并发限制 task asyncio.create_task(self._handle_task(task_id)) self._active_tasks.add(task) task.add_done_callback(self._active_tasks.discard) async def _handle_task(self, task_id: str): 任务处理核心逻辑 if not self._pool: raise RuntimeError(数据库连接池未初始化) async with self._sem: # 并发控制 async with self._pool.acquire() as conn: try: # 1. 执行业务逻辑 await self._execute_business_logic(task_id) # 2. 更新任务状态示例 print(fTask {task_id} completed successfully.) except Exception as e: print(fTask {task_id} failed: {e}) await self._log_failure(task_id, str(e)) async def _execute_business_logic(self, task_id: str): 模拟业务逻辑处理 print(fProcessing task {task_id}...) await asyncio.sleep(5) # 模拟耗时操作 print(fTask {task_id} completed.) async def _log_failure(self, task_id: str, error: str): 记录失败日志 print(fTask {task_id} failure logged: {error}) # 实际项目中可记录到数据库 async def stop(self): 优雅关闭 print(Shutting down gracefully...) # 等待正在运行的任务完成 if self._active_tasks: await asyncio.gather(*self._active_tasks, return_exceptionsTrue) # 清理资源 if self._listener_conn: await self._listener_conn.remove_listener(self._channel, self._on_notify) await self._listener_conn.close() if self._pool: await self._pool.close() print(Shutdown complete.) async def main(): worker TaskWorker(cfg.get_postgres_dsn(), cfg.postgres_channel()) loop asyncio.get_running_loop() stop_evt asyncio.Event() # 注册信号处理器优雅关闭 for sig in (signal.SIGINT, signal.SIGTERM): loop.add_signal_handler(sig, stop_evt.set) # 启动监听器 listen_task asyncio.create_task(worker.start()) # 等待停止事件 await stop_evt.wait() print(Shutdown signal received, stopping worker...) # 取消监听任务并等待完成 listen_task.cancel() await listen_task if __name__ __main__: asyncio.run(main())

本文来自互联网用户投稿,该文观点仅代表作者本人,不代表本站立场。本站仅提供信息存储空间服务,不拥有所有权,不承担相关法律责任。如若转载,请注明出处:http://www.coloradmin.cn/o/2497441.html

如若内容造成侵权/违法违规/事实不符,请联系多彩编程网进行投诉反馈,一经查实,立即删除!

相关文章

SpringBoot-17-MyBatis动态SQL标签之常用标签

文章目录 1 代码1.1 实体User.java1.2 接口UserMapper.java1.3 映射UserMapper.xml1.3.1 标签if1.3.2 标签if和where1.3.3 标签choose和when和otherwise1.4 UserController.java2 常用动态SQL标签2.1 标签set2.1.1 UserMapper.java2.1.2 UserMapper.xml2.1.3 UserController.ja…

wordpress后台更新后 前端没变化的解决方法

使用siteground主机的wordpress网站,会出现更新了网站内容和修改了php模板文件、js文件、css文件、图片文件后,网站没有变化的情况。 不熟悉siteground主机的新手,遇到这个问题,就很抓狂,明明是哪都没操作错误&#x…

网络编程(Modbus进阶)

思维导图 Modbus RTU(先学一点理论) 概念 Modbus RTU 是工业自动化领域 最广泛应用的串行通信协议,由 Modicon 公司(现施耐德电气)于 1979 年推出。它以 高效率、强健性、易实现的特点成为工业控制系统的通信标准。 包…

UE5 学习系列(二)用户操作界面及介绍

这篇博客是 UE5 学习系列博客的第二篇,在第一篇的基础上展开这篇内容。博客参考的 B 站视频资料和第一篇的链接如下: 【Note】:如果你已经完成安装等操作,可以只执行第一篇博客中 2. 新建一个空白游戏项目 章节操作,重…

IDEA运行Tomcat出现乱码问题解决汇总

最近正值期末周,有很多同学在写期末Java web作业时,运行tomcat出现乱码问题,经过多次解决与研究,我做了如下整理: 原因: IDEA本身编码与tomcat的编码与Windows编码不同导致,Windows 系统控制台…

利用最小二乘法找圆心和半径

#include <iostream> #include <vector> #include <cmath> #include <Eigen/Dense> // 需安装Eigen库用于矩阵运算 // 定义点结构 struct Point { double x, y; Point(double x_, double y_) : x(x_), y(y_) {} }; // 最小二乘法求圆心和半径 …

使用docker在3台服务器上搭建基于redis 6.x的一主两从三台均是哨兵模式

一、环境及版本说明 如果服务器已经安装了docker,则忽略此步骤,如果没有安装,则可以按照一下方式安装: 1. 在线安装(有互联网环境): 请看我这篇文章 传送阵>> 点我查看 2. 离线安装(内网环境):请看我这篇文章 传送阵>> 点我查看 说明&#xff1a;假设每台服务器已…

XML Group端口详解

在XML数据映射过程中&#xff0c;经常需要对数据进行分组聚合操作。例如&#xff0c;当处理包含多个物料明细的XML文件时&#xff0c;可能需要将相同物料号的明细归为一组&#xff0c;或对相同物料号的数量进行求和计算。传统实现方式通常需要编写脚本代码&#xff0c;增加了开…

LBE-LEX系列工业语音播放器|预警播报器|喇叭蜂鸣器的上位机配置操作说明

LBE-LEX系列工业语音播放器|预警播报器|喇叭蜂鸣器专为工业环境精心打造&#xff0c;完美适配AGV和无人叉车。同时&#xff0c;集成以太网与语音合成技术&#xff0c;为各类高级系统&#xff08;如MES、调度系统、库位管理、立库等&#xff09;提供高效便捷的语音交互体验。 L…

(LeetCode 每日一题) 3442. 奇偶频次间的最大差值 I (哈希、字符串)

题目&#xff1a;3442. 奇偶频次间的最大差值 I 思路 &#xff1a;哈希&#xff0c;时间复杂度0(n)。 用哈希表来记录每个字符串中字符的分布情况&#xff0c;哈希表这里用数组即可实现。 C版本&#xff1a; class Solution { public:int maxDifference(string s) {int a[26]…

【大模型RAG】拍照搜题技术架构速览:三层管道、两级检索、兜底大模型

摘要 拍照搜题系统采用“三层管道&#xff08;多模态 OCR → 语义检索 → 答案渲染&#xff09;、两级检索&#xff08;倒排 BM25 向量 HNSW&#xff09;并以大语言模型兜底”的整体框架&#xff1a; 多模态 OCR 层 将题目图片经过超分、去噪、倾斜校正后&#xff0c;分别用…

【Axure高保真原型】引导弹窗

今天和大家中分享引导弹窗的原型模板&#xff0c;载入页面后&#xff0c;会显示引导弹窗&#xff0c;适用于引导用户使用页面&#xff0c;点击完成后&#xff0c;会显示下一个引导弹窗&#xff0c;直至最后一个引导弹窗完成后进入首页。具体效果可以点击下方视频观看或打开下方…

接口测试中缓存处理策略

在接口测试中&#xff0c;缓存处理策略是一个关键环节&#xff0c;直接影响测试结果的准确性和可靠性。合理的缓存处理策略能够确保测试环境的一致性&#xff0c;避免因缓存数据导致的测试偏差。以下是接口测试中常见的缓存处理策略及其详细说明&#xff1a; 一、缓存处理的核…

龙虎榜——20250610

上证指数放量收阴线&#xff0c;个股多数下跌&#xff0c;盘中受消息影响大幅波动。 深证指数放量收阴线形成顶分型&#xff0c;指数短线有调整的需求&#xff0c;大概需要一两天。 2025年6月10日龙虎榜行业方向分析 1. 金融科技 代表标的&#xff1a;御银股份、雄帝科技 驱动…

观成科技:隐蔽隧道工具Ligolo-ng加密流量分析

1.工具介绍 Ligolo-ng是一款由go编写的高效隧道工具&#xff0c;该工具基于TUN接口实现其功能&#xff0c;利用反向TCP/TLS连接建立一条隐蔽的通信信道&#xff0c;支持使用Let’s Encrypt自动生成证书。Ligolo-ng的通信隐蔽性体现在其支持多种连接方式&#xff0c;适应复杂网…

铭豹扩展坞 USB转网口 突然无法识别解决方法

当 USB 转网口扩展坞在一台笔记本上无法识别,但在其他电脑上正常工作时,问题通常出在笔记本自身或其与扩展坞的兼容性上。以下是系统化的定位思路和排查步骤,帮助你快速找到故障原因: 背景: 一个M-pard(铭豹)扩展坞的网卡突然无法识别了,扩展出来的三个USB接口正常。…

未来机器人的大脑:如何用神经网络模拟器实现更智能的决策?

编辑&#xff1a;陈萍萍的公主一点人工一点智能 未来机器人的大脑&#xff1a;如何用神经网络模拟器实现更智能的决策&#xff1f;RWM通过双自回归机制有效解决了复合误差、部分可观测性和随机动力学等关键挑战&#xff0c;在不依赖领域特定归纳偏见的条件下实现了卓越的预测准…

Linux应用开发之网络套接字编程(实例篇)

服务端与客户端单连接 服务端代码 #include <sys/socket.h> #include <sys/types.h> #include <netinet/in.h> #include <stdio.h> #include <stdlib.h> #include <string.h> #include <arpa/inet.h> #include <pthread.h> …

华为云AI开发平台ModelArts

华为云ModelArts&#xff1a;重塑AI开发流程的“智能引擎”与“创新加速器”&#xff01; 在人工智能浪潮席卷全球的2025年&#xff0c;企业拥抱AI的意愿空前高涨&#xff0c;但技术门槛高、流程复杂、资源投入巨大的现实&#xff0c;却让许多创新构想止步于实验室。数据科学家…

深度学习在微纳光子学中的应用

深度学习在微纳光子学中的主要应用方向 深度学习与微纳光子学的结合主要集中在以下几个方向&#xff1a; 逆向设计 通过神经网络快速预测微纳结构的光学响应&#xff0c;替代传统耗时的数值模拟方法。例如设计超表面、光子晶体等结构。 特征提取与优化 从复杂的光学数据中自…