Intermittent Telegram Node Issues in n8n: Seeking Insights from the Community

[NickSwiss2024]

As an n8n user, I've successfully implemented workflows that integrate with Telegram via the Telegram node. These workflows power three different AI-driven projects that I rely on daily. While the integration generally works well, I've encountered a recurring issue that's challenging the reliability of my system.

The problem is specific and perplexing: one of my Telegram nodes occasionally stops functioning after several days or weeks of flawless operation. The other Telegram nodes for my other projects continue to run without any problems. When this issue occurs, I’ve found that simply deactivating and reactivating the affected workflow resolves it immediately. However, this workaround isn’t ideal for a production-grade setup, especially when delivering solutions to real clients.

Here are some key observations about this issue:

Scope of the Problem: Only one Telegram node fails at a time, and the failure occurs sporadically without a clear pattern.
Reliability Elsewhere: The rest of the Telegram nodes and other input mechanisms, such as the Chat node, operate reliably for weeks or even months.
Temporary Fix: Deactivating and reactivating the affected workflow resolves the issue instantly, but this manual intervention is not sustainable for professional use.
I've considered implementing an automated script to periodically deactivate and reactivate the workflows as a stopgap solution. However, this feels like a band-aid rather than a proper fix. I suspect there may be a deeper issue with the Telegram node’s connection or session handling.

My Questions for the Community:

Have you experienced similar intermittent failures with the Telegram node in n8n?
What strategies have you used to maintain long-term reliability for Telegram integrations?
Is there a known root cause or fix for this type of issue?
Are there alternative ways to monitor and handle workflow failures more gracefully within n8n?
This problem is critical for me, as I plan to scale these AI-driven Telegram bots for clients. Any insights, advice, or shared experiences from fellow n8n users would be greatly appreciated. Let’s collaborate to make n8n even more robust for production-grade use cases!

I am using the cloud hosted version.

core

• n8nVersion: 1.75.2
• platform: npm
• nodeJsVersion: 20.18.1
• database: sqlite
• executionMode: regular
• concurrency: 5
• license: community

storage

• success: all
• error: all
• progress: false
• manual: true
• binaryMode: filesystem

pruning

• enabled: true
• maxAge: 168 hours
• maxCount: 2500 executions

client

• userAgent: mozilla/5.0 (x11; linux x86_64) applewebkit/537.36 (khtml, like gecko) chrome/132.0.0.0 safari/537.36
• isTouchDevice: false

Generated at: 2025-01-29T20:15:56.954Z}

[Joffcom]

Hey @NickSwiss2024,

We have created an internal ticket to look into this which we will be tracking as "GHC-664"

[Joffcom]

Hey @NickSwiss2024

You can find our community on our forum and on Discord, We use GitHub issues just for tracking bugs.

One thing to remember with Telegram is that each bot can only have one webhook registered, one of the problems I have seen is users testing an active workflow which causes the webhook url to be overwritten in Telegram and often needs the workflow to be deactivated and activated again so it registers. Is it possible that this is what you are seeing?