How to use Amazon documentdb with langchain?

Question

I know documentdb isn't officialy or community supported yet by langchain but am trying to use their MongoDb atlas one kinda like this question but my problem is am getting this error :

pymongo.errors.ServerSelectionTimeoutError: docdb-restofit.cluster-cpm2mmw0q9qb.us-east-1.docdb.amazonaws.com:27017: timed out (configured timeouts: socketTimeoutMS: 20000.0ms, connectTimeoutMS: 20000.0ms)

I am fairly new to aws and documentdb and cant seem to find a solution will really appreciate the help.

Here is what I have done till now to accomplish this, I am new to aws so a little confused.

Documentdb, my steps:

1. Created a new cluster with name and password, everything else is default
2. switched off tls connection, since I wanted a quick and easy way to connect to my cluster
3. use the mongodb uri

Now my code for langchain

from dotenv import load_dotenv
from langchain.document_loaders import TextLoader
from langchain.embeddings import OpenAIEmbeddings
from langchain.text_splitter import CharacterTextSplitter
from langchain_community.document_loaders import PyPDFLoader
from langchain_community.vectorstores import Chroma, MongoDBAtlasVectorSearch
from pymongo import MongoClient

load_dotenv()

text_splitter = CharacterTextSplitter(
    separator="\n",
    chunk_size=500,
)

emb = OpenAIEmbeddings()


loader = PyPDFLoader("./101.pdf")
pages = loader.load_and_split()
docs = loader.load_and_split(
    text_splitter=text_splitter,
)


client = MongoClient()

DB_NAME = "langchain_db"
COLLECTION_NAME = "test"
ATLAS_VECTOR_SEARCH_INDEX_NAME = "index_name"

MONGODB_COLLECTION = client[DB_NAME][COLLECTION_NAME]

vector_search = MongoDBAtlasVectorSearch.from_documents(
    documents=docs,
    embedding=OpenAIEmbeddings(disallowed_special=()),
    collection=MONGODB_COLLECTION,
    index_name=ATLAS_VECTOR_SEARCH_INDEX_NAME,
)

Answer 1

I solved it and it was due to VPC policy of amazon aws this was helpful