N8n may have run out of memory while running this execution

Hi,

I am having troubles with my Ai Agents in the recent days. Lots of them get this error:
“Execution stopped at this node
n8n may have run out of memory while running this execution. More context and tips on how to avoid this in the docs

I saw the docs and it says that i can do 2 things, increase the memory, by upgrading the plan or decrease the flow memory.

But none of these work for me. I have the Starter pack and i am using around 400 to 500 executions monthly. Doesn’t make any sense to go for 10k.

Also my flow worked always, i had hounders of working executions till now and never had these issues, just the recent days lots of my executions get stopped because of this.

the agent has a simple prompt that adds a youtube channel in my db if it doesn’t exists, I can’t understand why it stops. It also stops to other workflows that in the past were working great.

I wanted to use the cloud as it’s easier to manage, i am happy to pay for the n8n could subscription if i don’t stress about servers and unavailability of the servers. But in the recent days the n8n cloud seems unstable.

Can you please help me out to figure this out?

The Agent:

{
“nodes”: [
{
“parameters”: {
“model”: “openai/gpt-4o-mini”,
“options”: {
“temperature”: 0
}
},
“type”: “@n8n/n8n-nodes-langchain.lmChatOpenRouter”,
“typeVersion”: 1,
“position”: [
-100,
610
],
“id”: “02cb63b1-074b-4c85-99d2-a1dc57a50698”,
“name”: “OpenRouter Chat Model”,
“credentials”: {
“openRouterApi”: {
“id”: “ZzfQwzxdwzfe1J2i”,
“name”: “Mihai Pol :: OpenRouter”
}
}
},
{
“parameters”: {
“resource”: “databasePage”,
“operation”: “getAll”,
“databaseId”: {
“__rl”: true,
“value”: “214444c6-bc17-8018-85a4-ff4601ae6b39”,
“mode”: “list”,
“cachedResultName”: “Youtube Channels”,
“cachedResultUrl”: “Notion
},
“limit”: 1,
“filterType”: “manual”,
“filters”: {
“conditions”: [
{
“key”: “Youtube ID|rich_text”,
“condition”: “equals”,
“richTextValue”: “={{ $(‘Video Details’).first().json.channelId }}”
}
]
},
“options”: {}
},
“type”: “n8n-nodes-base.notionTool”,
“typeVersion”: 2.2,
“position”: [
20,
610
],
“id”: “98f31564-2bde-44dd-a174-10f9c4226628”,
“name”: “Check If Channel Exists1”,
“credentials”: {
“notionApi”: {
“id”: “S8yP0qhSpQQeo8qB”,
“name”: “Mihai Pol :: Notion”
}
}
},
{
“parameters”: {
“resource”: “databasePage”,
“databaseId”: {
“__rl”: true,
“value”: “214444c6-bc17-8018-85a4-ff4601ae6b39”,
“mode”: “list”,
“cachedResultName”: “Youtube Channels”,
“cachedResultUrl”: “Notion
},
“title”: “={{ $(‘Video Details’).first().json.channelName }}”,
“propertiesUi”: {
“propertyValues”: [
{
“key”: “Youtube ID|rich_text”,
“textContent”: “={{ $(‘Video Details’).first().json.channelId }}”
},
{
“key”: “URL|url”,
“urlValue”: “={{ $(‘Video Details’).first().json.channelUrl }}”
}
]
},
“options”: {
“iconType”: “file”,
“icon”: “={{ /n8n-auto-generated-fromAI-override/ $fromAI(‘Icon’, ``, ‘string’) }}”
}
},
“type”: “n8n-nodes-base.notionTool”,
“typeVersion”: 2.2,
“position”: [
140,
610
],
“id”: “fb05ed48-4a12-45d4-b1a7-263f8cb28c44”,
“name”: “CreateChannel”,
“credentials”: {
“notionApi”: {
“id”: “S8yP0qhSpQQeo8qB”,
“name”: “Mihai Pol :: Notion”
}
}
},
{
“parameters”: {
“jsonSchemaExample”: “{\n\t"notionChannelId": "217444c6-bc17-812c-8193-cc4974bdc6eb"\n}”
},
“type”: “@n8n/n8n-nodes-langchain.outputParserStructured”,
“typeVersion”: 1.2,
“position”: [
380,
610
],
“id”: “38016783-40e9-4d2a-a5a4-89c33e2eb848”,
“name”: “Structured Output Parser1”
},
{
“parameters”: {
“toolDescription”: “Get Channel Thumbnail URL”,
“method”: “POST”,
“url”: “https://api.apify.com/v2/acts/streamers~youtube-channel-scraper/run-sync-get-dataset-items?token=apify_api_eZVNYR2bqKWIjdVO44CvkZcm20jYfK49Rudt”,
“sendBody”: true,
“specifyBody”: “json”,
“jsonBody”: “={\n "details": false,\n "max_comments": 1,\n "max_videos": 1,\n "scrap_captions": false,\n "startUrls": [\n {\n "url": "{{ $(‘Video Details’).first().json.channelUrl }}",\n "method": "GET"\n }\n ]\n}”,
“options”: {}
},
“type”: “n8n-nodes-base.httpRequestTool”,
“typeVersion”: 4.2,
“position”: [
260,
610
],
“id”: “383c6ed7-f5aa-4ce9-bff6-83facb2b8b3a”,
“name”: “Get Channel Thumbnail”
},
{
“parameters”: {
“promptType”: “define”,
“text”: “=youtubeChannelId: {{ $(‘Video Details’).first().json.channelId }}”,
“hasOutputParser”: true,
“options”: {
“systemMessage”: “=You are a Notion CRUD AI Expert Agent. Your job is to Check If Channel Exists and create it if not.\n\nTOOLS: USE THEM EXACTLY IN THIS ORDER\n1 checkIfChannelExists - use this to check if youtube channel exists\n2 getChannelthumbnail - use this only if checkIfChannelExists returned empty response and send thumbnailUrl to next createChannel\n3 createChannel - use this only if checkIfChannelExists returned empty response\n\nOUTPUT:\n{\n "notionChannelId": "217444c6-bc17-812c-8193-cc4974bdc6eb"\n}”
}
},
“type”: “@n8n/n8n-nodes-langchain.agent”,
“typeVersion”: 2,
“position”: [
60,
400
],
“id”: “e7a04545-8faf-4491-9a30-286f905553ee”,
“name”: “Youtube Channels Agent”
}
],
“connections”: {
“OpenRouter Chat Model”: {
“ai_languageModel”: [
[
{
“node”: “Youtube Channels Agent”,
“type”: “ai_languageModel”,
“index”: 0
}
]
]
},
“Check If Channel Exists1”: {
“ai_tool”: [
[
{
“node”: “Youtube Channels Agent”,
“type”: “ai_tool”,
“index”: 0
}
]
]
},
“CreateChannel”: {
“ai_tool”: [
[
{
“node”: “Youtube Channels Agent”,
“type”: “ai_tool”,
“index”: 0
}
]
]
},
“Structured Output Parser1”: {
“ai_outputParser”: [
[
{
“node”: “Youtube Channels Agent”,
“type”: “ai_outputParser”,
“index”: 0
}
]
]
},
“Get Channel Thumbnail”: {
“ai_tool”: [
[
{
“node”: “Youtube Channels Agent”,
“type”: “ai_tool”,
“index”: 0
}
]
]
},
“Youtube Channels Agent”: {
“main”: [

]
}
},
“pinData”: {},
“meta”: {
“instanceId”: “3b4227a77ae543d4e30b89a51cbe6af77651dde3896ed1a5515a746277fbfaec”
}
}

Information on your n8n setup

  • n8n version: 1.10
  • Database (default: SQLite): notion
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • Running n8n via (Docker, npm, n8n cloud, desktop app): n8n cloud
  • Operating system: macos

Dear @Mihai_Pol that may be the update issue.

Recently n8n has carry out new update.

I think if you update the n8n the issue will solve out.

Let;s try it and tell me whether it works or not.

Happy to help…

I believe the n8n in this case is in cloud. Based on

Dear @jabbson How’s everything going on. I am learning a lot from you and the stuff that you are sharing. Thank you so much for your kind efforts.

Recently n8n update some changes into cloud one too, and suggest us to update the version of n8n.

So, I am suggesting that one.

By the way, how it feel to sit on the top of the lead board ??

Thank you so much for your kind efforts again…

Ah, I didn’t know you have to update the cloud version too, thought it happens automagically. See, I am learning something too :slight_smile:

That’s funny indeed @jabbson

see this

1 Like

You can also, execute subworkflows inside your main workflow this helps keeps memory low and clean after sub workflow execute it cleans, so this may help in your situation as when calling subworkflow it’s still only classed as 1 execution the first one which runs it, even if you call multiple subworkflows.

Hope that makes sense.

Also you mentioned about storing in DB, how large is the db being called? Maybe it’s grown so much overtime each call now pulls alot of data?

Samuel