Documentation/Default Script to be updated

[Multidivide]

Description

The Worker.py file points to a localhost port (in the config file it links to) that I think has been deprecated, when i leave it as the default port 8080 it provides an error of the API being offline. However if I point it to port 4000 it appears to be functioning correctly as it is now waiting for a connector.

Additonally, RabbitMQ requires the management plugin as the worker needs to be able to interact with the rabbitMQ port 15672, the following command is required before worker.py is executed.

Environment

1. OS (where OpenCTI server runs): Ubuntu 18.04
2. OpenCTI version: 2.0.2
3. OpenCTI client: frontend

Actual Output

Below were some of the error messages I've received;

1. When running the worker by default
   Traceback (most recent call last): File "worker.py", line 173, in <module> worker = Worker() File "worker.py", line 119, in __init__ self.api = OpenCTIApiClient(self.opencti_url, self.opencti_token) File "/home/opencti/.local/lib/python3.6/site-packages/pycti/api/opencti_api_client.py", line 52, in __init__ raise ValueError('OpenCTI API seems down') ValueError: OpenCTI API seems down

2. When running on port 4000, but then struggles with RabbitMQ as a management plugin is missing, error log is from the running OpenCTI Server;
   error: [OPENCTI] Technical error > connect ECONNREFUSED 127.0.0.1:15672 {"locations":[{"line":8,"column":3}],"path":["rabbitMQMetrics"],"extensions":{"code":"INTERNAL_SERVER_ERROR","exception":{"config":{"url":"http://localhost:15672/api/overview","method":"get","headers":{"Accept":"application/json, text/plain, */*","User-Agent":"axios/0.19.0"},"auth":{"username":"guest","password":"guest"},"baseURL":"http://localhost:15672","transformRequest":[null],"transformResponse":[null],"timeout":0,"withCredentials":true,"xsrfCookieName":"XSRF-TOKEN","xsrfHeaderName":"X-XSRF-TOKEN","maxContentLength":-1},"code":"ECONNREFUSED"}}}

Solution

1. Change the config file to localhost:4000
2. rabbitmq-plugins enable rabbitmq_management before you run the worker.py script

Outcome

Next steps I think would be just including these in the documentation, I worked it out using the error codes but wanted to share in case others had the problem.

[graememeyer]

Seems to still be an issue on 2.1.4

If you're using the test configuration (docker-compose_test.yml), you can use these shell commands to fix:

sed -i 's/8080/4000/g' docker-compose.yml
sed -i 's/ChangeMeAccess/NewAccessINSERTRANDOMSTRINGHERE/g' docker-compose.yml
sed -i 's/ChangeMeKey/NewAccessKeyINSERTRANDOMSTRINGHERE/g' docker-compose.yml