Error: Cannot read properties of undefined (reading 'client')

Questions
1

Hi everyone,

I’m running into the following error when using a MongoDB node as agent memory in my workflow:

Error: Cannot read properties of undefined (reading 'client')

NodeOperationError: Error in sub-node MongoDB Chat Memory at ExecuteContext.getInputConnectionData (C:\Users\FritzerJos\AppData\Roaming\npm\node_modules\n8n\node_modules\n8n-core\src\execution-engine\node-execution-context\utils\get-input-connection-data.ts:352:11) at processTicksAndRejections (node:internal/process/task_queues:105:5) at ExecuteContext.getInputConnectionData (C:\Users\FritzerJos\AppData\Roaming\npm\node_modules\n8n\node_modules\n8n-core\src\execution-engine\node-execution-context\execute-context.ts:180:10) at getOptionalMemory (C:\Users\FritzerJos\AppData\Roaming\npm\node_modules\n8n\node_modules@n8n\n8n-nodes-langchain\nodes\agents\Agent\agents\ToolsAgent\common.ts:314:10) at ExecuteContext.toolsAgentExecute (C:\Users\FritzerJos\AppData\Roaming\npm\node_modules\n8n\node_modules@n8n\n8n-nodes-langchain\nodes\agents\Agent\agents\ToolsAgent\V2\execute.ts:194:17) at ExecuteContext.execute (C:\Users\FritzerJos\AppData\Roaming\npm\node_modules\n8n\node_modules@n8n\n8n-nodes-langchain\nodes\agents\Agent\V2\AgentV2.node.ts:131:10) at WorkflowExecute.executeNode (C:\Users\FritzerJos\AppData\Roaming\npm\node_modules\n8n\node_modules\n8n-core\src\execution-engine\workflow-execute.ts:1093:8) at WorkflowExecute.runNode (C:\Users\FritzerJos\AppData\Roaming\npm\node_modules\n8n\node_modules\n8n-core\src\execution-engine\workflow-execute.ts:1274:11) at C:\Users\FritzerJos\AppData\Roaming\npm\node_modules\n8n\node_modules\n8n-core\src\execution-engine\workflow-execute.ts:1699:27 at C:\Users\FritzerJos\AppData\Roaming\npm\node_modules\n8n\node_modules\n8n-core\src\execution-engine\workflow-execute.ts:2315:11

I’m on n8n version 1.116.2.

What I’ve checked so far:

  • The MongoDB connection test in the node’s credentials setup passes successfully.

  • The connection details (URI, authentication, etc.) are correct and match my MongoDB instance.

Additional context:

  • The error occurs specifically when the workflow tries to use the MongoDB node for agent memory operations.

  • No issues with other nodes or database connections.

Has anyone encountered this issue or have suggestions on how to resolve it? Any insights would be greatly appreciated!

Thanks in advance!

2 Likes
2

I have the same issue after updating from 1.111.1 to 1.114.4

3

And after downgrading to 1.111.1 again, still not works. Remains broken

4

I have this problem after update version . I return version back but problem not solve

5

Recently, I also faced the same error. How can I solve it?

6

Hey, I think this could be the issue GitHub · Where software is built

1 Like
7

I had an issue of Error: Cannot read properties of undefined (reading ‘client’) in n8n Version 1.118.1, and I solved it by adding this code to: C:\Users\AppData\Roaming\npm\node_modules\n8n\node_modules@n8n\n8n-nodes-langchain\dist\nodes\vector_store\VectorStoreMongoDBAtlas\VectorStoreMongoDBAtlas.node.js

      if (collection) {
        if (collection.db === void 0) {
          collection.db = db;
        }
        if (collection.db?.client && typeof collection.db.client.appendMetadata !== "function") {
          collection.db.client.appendMetadata = () => {
          };
        }

8

Where did you put this code?

9

Find:

const collection = db.collection(collectionName);
const indexes = await collection.listSearchIndexes().toArray();
const indexExists = indexes.some((index) => index.name === mongoVectorIndexName);

Replace with:

const collection = db.collection(collectionName);

if (collection) {
  if (collection.db === void 0) {
    collection.db = db;
  }
  if (collection.db && collection.db.client && typeof collection.db.client.appendMetadata !== "function") {
    collection.db.client.appendMetadata = () => {};
  }
}

const indexes = await collection.listSearchIndexes().toArray();
const indexExists = indexes.some((index) => index.name === mongoVectorIndexName);

Then find:

const collection = db.collection(collectionName);
await ExtendedMongoDBAtlasVectorSearch.fromDocuments(documents, embeddings, {
  collection,
  indexName: mongoVectorIndexName,
  // Default index name
  textKey: metadataFieldName,
  // Field containing raw text
  embeddingKey: embeddingFieldName
  // Field containing embeddings
});

replace with:

const collection = db.collection(collectionName);

if (collection) {
  if (collection.db === void 0) {
    collection.db = db;
  }
  if (collection.db && collection.db.client && typeof collection.db.client.appendMetadata !== "function") {
    collection.db.client.appendMetadata = () => {};
  }
}

await ExtendedMongoDBAtlasVectorSearch.fromDocuments(documents, embeddings, {
  collection,
  indexName: mongoVectorIndexName,
  // Default index name
  textKey: metadataFieldName,
  // Field containing raw text
  embeddingKey: embeddingFieldName
  // Field containing embeddings
});

Worked for me :slight_smile: