我正在考慮開發一個小工具,用來監控你的自架 n8n 工作流程,並在執行失敗或卡住時立即通知你(Discord/Telegram/電子郵件)— 因為 healthchecks.io/Cronitor 無法理解 n8n 的語義(哪個節點出問題了、哪個工作流程出問題了)。這對你有用嗎?你目前在監控方面支付多少費用(如果有的話)?
嘿 @Thomas42-maker,在等待回覆的同時,以下是一些可能有幫助的資源:
@achamm、@Emmas - 你們之前幫助解決過類似的問題,可以看一下嗎?
由 n8n 社群機器人自動建議。這是試點版本 - 請在此分享反饋。
歡迎 @Thomas42-maker!
是的,這會很有用 - 大多數人把 n8n 內建的錯誤觸發工作流程與其他東西拼湊在一起,因為它只在硬體故障時觸發,而不是工作流程內的掛起或部分/靜默故障。
我希望專門構建的工具能涵蓋的幾個差距:
- 按節點的故障上下文(哪個節點出錯,包含實際的錯誤負載),而不只是「工作流程 X 失敗」
- 檢測長時間運行/卡住的執行,而不只是崩潰的執行 - 工作流程卡在「執行中」一小時通常意味著 webhook 從未解決或外部 API 掛起
- 區分預期的錯誤(例如你已經重試的 Slack 速率限制)和真實錯誤,這樣你就不會被噪音所困擾
目前大多數自託管用戶我看到的都是手動將錯誤觸發器連接到 Slack/Telegram 節點就完事,但一旦你有 20 多個工作流程就會崩潰,因為你會失去按節點的可見性。如果你的工具能直接讀取執行數據(通過 API 或數據庫)並將節點級故障與工作流程上下文關聯起來,僅這一點就能勝過自己開發的方案。
Thanks, this is exactly the kind of detail I needed. Quick follow-up: would you (or people you’ve seen) actually pay for this as a subscription, or is it something you’d only use if it were free/open-source? And roughly how many workflows does someone need before this pain becomes serious enough to pay for?
From what I’ve seen, most self-hosted users won’t pay until they hit 15-20+ production workflows, that’s usually when manual monitoring stops scaling and a missed failure actually costs money or trust with a client. Below that, people lean on free options (Error Trigger + Slack, or Uptime Kuma push) since the pain isn’t big enough yet. On pricing, a low-cost subscription (10-20 USD/month) beats one-time or open-source for this kind of tool, because ongoing alerting/monitoring feels like infrastructure people expect to pay recurring for, similar to Uptime Kuma Cloud or Better Stack. I’d focus your free tier on catching people right before that 15-20 workflow threshold, then convert them once they feel the pain.