Issue with Closing Node Detail Page in n8n Workflow

Questions
1

Hi,

I encountered a serious issue while renaming a filter node using a Chinese phonetic input method. After pressing enter, it automatically opened the filter node’s detail page, and now I am unable to close the current window (node detail page).

Regardless of refreshing the page, navigating back and reopening the workflow, or restarting my self-deployed n8n instance, as soon as I enter this workflow page, I am stuck on the current window with no ability to perform any actions. Even the “back to canvas” button in the top left corner is unresponsive.

I have recorded a screen capture of the issue using CleanShotX: CleanShot 2024-05-24 at 00.47.18 · CleanShot Cloud

Could you please advise on how to resolve this? I desperately need support as this workflow contains a lot of important settings.

Thank you!

Information on your n8n setup

  • n8n version: 1.42.1
  • Database (default: SQLite): Post greSQL
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • Running n8n via (Docker, npm, n8n cloud, desktop app): zeabur
  • Operating system:
2

Additionally, even if I duplicate the workflow and enter the duplicated version, the same issue occurs. This confirms that it is likely a bug in n8n itself. Here is another screen recording showing the issue: CleanShot 2024-05-24 at 01.01.50 · CleanShot Cloud

Console errors include:

Unchecked runtime.lastError: The message port closed before a response was received.
inject.js:1 [Deprecation] Listener added for a 'DOMNodeInserted' mutation event. This event type is deprecated, and will be removed from this browser very soon. Usage of this event listener will cause performance issues today, and represents a large risk of future site breakage. Consider using MutationObserver instead. See https://chromestatus.com/feature/5083947249172480 for more information.
NodeView-8Nd44_r7.js:8 TypeError: Cannot read properties of null (reading 'id')
    at Proxy.node (NodeView-8Nd44_r7.js:8:12830)
    at Ze (vendor-KLOX9vbo.js:13:1373)
    at ke (vendor-KLOX9vbo.js:13:1443)
    at I (vendor-KLOX9vbo.js:13:12595)
    at fo (vendor-KLOX9vbo.js:13:2672)
    at X (vendor-KLOX9vbo.js:13:42496)
    at Xt.A [as fn] (vendor-KLOX9vbo.js:13:41510)
    at Xt.run (vendor-KLOX9vbo.js:9:1517)
    at d.update (vendor-KLOX9vbo.js:13:42352)
    at ce (vendor-KLOX9vbo.js:13:41254)
3

Have you tried to clear the cookies? And is the issue persists if you rename the filter node and set some name?

Plus can you share the workflow itself?

4

Update: I found a workaround to bypass this bug!

By using the browser’s Inspect tool, I manually deleted the HTML elements blocking the UI. Specifically, I removed the
<div class="el-overlay" style="z-index: 2009;">...</div>
element as shown in the attached screenshot. This allowed me to interact with the workflow and rename the filter node, resolving the issue.

CleanShot 2024-05-24 at 10.03.37@2x
CleanShot 2024-05-24 at 10.03.37@2x1920×1080 80.1 KB

It seems that if the node name is blank, it causes the UI to get stuck. This might be a significant bug.

Thank you for your assistance!

1 Like