How do I connect to AWS Instance of Opensearch using OpenSearchDocumentStore

[gooseillo]

This is what I am trying. The opensearch instance has http auth and is currently open to the internet.

from aws_requests_auth.aws_auth import AWSRequestsAuth

aws_auth = AWSRequestsAuth(
aws_access_key='xxxxxx',
aws_secret_access_key='xxxxxx',
aws_host='xxxxxx.es.amazonaws.com',
aws_region='us-east-2',
aws_service='es'
)

from haystack.document_stores import OpenSearchDocumentStore

document_store = OpenSearchDocumentStore(
scheme="https",
host="xxxxxxxx.us-east-2.es.amazonaws.com",
username='xxxxx',
password='xxxxx',
index="xxxxx",
embedding_dim=1024,
similarity="cosine",
verify_certs=True,
aws4auth = aws_auth
)

Error message

ConnectionError: Initial connection to Opensearch failed with error 'ConnectionError((, 'Connection toxxxx.us-east-2.es.amazonaws.com timed out. (connect timeout=30)')) caused by: ConnectTimeoutError((, 'Connection to xxxx.us-east-2.es.amazonaws.com timed out. (connect timeout=30)'))'
Make sure an Opensearch instance is running at `xxxxus-east-2.es.amazonaws.com` and that it has finished booting (can take > 30s).

Expected behavior
It should connect. I can connect to the instance using opensearchpy

Additional context
Name: farm-haystack
Version: 1.14.0

Name: opensearch-py
Version: 2.2.0

How do I connect to the AWS Opensearch instance so I can index and embedd documents?
Is there also al way to create the embedding locally and the so a snapshot and restore to the index?

[ZanSara]

Hey @gooseillo are you able to connect to that opensearch instance outside of haystack? If yes, could you upgrade Haystack to the latest release and let us know if the problem persists?