Neo4j instructions have conflicting information

[arun-gupta]

Following the instructions at https://github.com/opea-project/GenAIComps/tree/main/comps/knowledgegraphs/langchain

The instructions require Neo4J to be running. Then trying to start neo4j docker image gives the following error:

ubuntu@ip-172-31-57-26:~$ sudo docker run --rm     --publish=7474:7474 --publish=7687:7687     --env NEO4J_AUTH=$NEO4J_USER/$NEO4J_PASSWORD     --volume=$PWD/neo4j_data:"/data"     --env='NEO4JLABS_PLUGINS=["apoc"]'     neo4j
docker: Error response from daemon: driver failed programming external connectivity on endpoint elastic_chatterjee (b4afd0278d9f49a0532b6296b2231c70412560c08758a77714e9b24db4f34e32): failed to bind port 0.0.0.0:7474/tcp: Error starting userland proxy: listen tcp4 0.0.0.0:7474: bind: address already in use.

To keep it cloud native, we should only run it using Docker image.

Trying to run only the docker image gives the following error:

buntu@ip-172-31-57-26:~$ sudo docker run --rm     --publish=7474:7474 --publish=7687:7687     --env NEO4J_AUTH=$NEO4J_USER/$NEO4J_PASSWORD     --volume=$PWD/neo4j_data:"/data"     --env='NEO4JLABS_PLUGINS=["apoc"]'     neo4j
NEO4JLABS_PLUGINS has been renamed to NEO4J_PLUGINS since Neo4j 5.0.0.
The old name will still work, but is likely to be deprecated in future releases.
Installing Plugin 'apoc' from /var/lib/neo4j/labs/apoc-*-core.jar to /var/lib/neo4j/plugins/apoc.jar
Applying default values for plugin apoc to neo4j.conf
/neo4j is invalid
Invalid value for NEO4J_AUTH: '/neo4j'

[XinyaoWa]

(1) The first error only comes address conflict, means there is another container running on the same address, you can kill it first. (2) For the second error, please set a "NEO4J_PASSWORD" first with "export NEO4J_PASSWORD=${define_a_password}". (3) The link you follow will not be maintained later, you can follow https://github.com/opea-project/GenAIComps/tree/main/comps/dataprep/neo4j/langchain, which is the new guideline, and any questions about knowledgegraph can connect @XuhuiRen

[arun-gupta]

@XuhuiRen mentioned the new set of instructions is missing installing opea-comps. Followed this new set of instructions and now getting:

ubuntu@ip-172-31-55-237:~/GenAIComps/comps/dataprep/neo4j/langchain$ python3 prepare_doc_neo4j.py
/home/ubuntu/.local/lib/python3.12/site-packages/pydantic/_internal/_fields.py:132: UserWarning: Field "model_name_or_path" in Audio2TextDoc has conflict with protected namespace "model_".

You may be able to resolve this warning by setting `model_config['protected_namespaces'] = ()`.
  warnings.warn(
Traceback (most recent call last):
  File "/home/ubuntu/GenAIComps/comps/dataprep/neo4j/langchain/prepare_doc_neo4j.py", line 21, in <module>
    from comps.dataprep.utils import (
ModuleNotFoundError: No module named 'comps.dataprep'

Here is the package list:

ubuntu@ip-172-31-61-212:~/GenAIComps/comps/dataprep/neo4j/langchain$ pip3 list | grep comps
opea-comps         

Here are my exact instructions: https://gist.github.com/arun-gupta/d08a191d495b544a791cac0eefacd334