louislam/uptime-kuma

Notification lost if service is temporarily unavailable

Open

#6,046 创建于 2025年8月8日

在 GitHub 查看
 (4 评论) (0 反应) (0 负责人)JavaScript (86,760 star) (7,844 fork)batch import
A:notificationsfeature-requesthelp wanted

描述

📑 I have found these related issues/pull requests

Searched but didn't find any related issue/PR

🛡️ Security Policy

📝 Description

If the notification server is temporarily unreachable/unavailable at the moment of the down/up event, no retries are attempted to send the notification. I have a dual internet connection with a primary and a backup link. If the primary link fails there is a small delay (seconds) before the primary router switches to the backup link and during this short period of time the Telegram server is inaccessible from Uptime Kuma. The down notification of the primary link is never received even if Telegram is reachable just few seconds later. I have not checked if this is just related to Telegram or a more general behavior also affecting other notification channels.

👟 Reproduction steps

  1. disable internet connection (or just api.telegram.org) from Uptime Kuma
  2. down any monitor configured with a Telegram notification
  3. wait 5-10 seconds
  4. re-enable internet/telegram connectivity
  5. check that no down notification is sent

👀 Expected behavior

The down notification is retried and sent as soon as a connection with the notification service is available.

😓 Actual Behavior

No notification is sent

🐻 Uptime-Kuma Version

1.23.15

💻 Operating System and Arch

Debian GNU/Linux 12.8 (bookworm)

🌐 Browser

Firefox 141.0.3

🖥️ Deployment Environment

  • Runtime Environment:
    • Docker: Docker version 27.3.1, build ce12230
    • Docker Compose: Docker Compose version v2.29.7
  • Database:
    • SQLite: Embedded
  • Database Storage:
    • Filesystem:
      • Linux: Btrfs
    • Storage Medium: NVMe + SSD (RAID1)
  • Uptime Kuma Setup:
    • Number of monitors: 22

📝 Relevant log output

贡献者指南

技术栈
javascriptnodejssqlite
领域
backendinfrastructure
议题类型
bug
难度面向新贡献者的预计实现难度，1 表示很小改动，5 表示专家级工作。
3
预计时间有经验贡献者完成调查、实现、测试并准备 pull request 的粗略时间范围。
half day
活动状态议题当前的可参与程度：新鲜、活跃、陈旧、阻塞或等待维护者输入。
fresh
清晰度议题是否清楚说明期望改动、验收标准和下一步。
clear
前置要求
Node.js knowledgeAsync/await patternsNotification system overview
新手友好度1-100 的估计分数，表示该议题对首次贡献者的友好程度。
70
研究方向
Investigate the notification sending logic in Uptime Kuma, likely in 'server/notification.js' or similar files. Identify where the failure occurs when the notification server is unreachable. Implement a retry mechanism with exponential backoff or a notification queue. Consider adding a persistent queue (e.g., using SQLite) to store failed notifications and retry them periodically. Check for any existing event loops or schedulers that could trigger retries. Ensure no duplicate notifications are sent after the outage resolves.