External 'requests.post' not working about machine-learning HOT 8 OPEN

The above 500 error is achieved when the correct port rules are configured to be open. Otherwise, variations of different http errors are returned. However, the former http error indicates either there was a problem with the nginx-api reverse proxy, or the corresponding gunicorn-api webserver.

from machine-learning.

Part of our earlier confusion was due to the fact that the reverse proxy for the backend api, was configured on port 9595. However, we were making a request the frontend api at port 8585

from machine-learning.

With the adjusted port 9595, as well as using token = login.json(), we receive a timeout:

root@ubuntu-xenial:/vagrant# python3 test.py
Traceback (most recent call last):
  File "/usr/lib/python3/dist-packages/urllib3/connection.py", line 137, in _new_conn
    (self.host, self.port), self.timeout, **extra_kw)
  File "/usr/lib/python3/dist-packages/urllib3/util/connection.py", line 91, in create_connection
    raise err
  File "/usr/lib/python3/dist-packages/urllib3/util/connection.py", line 81, in create_connection
    sock.connect(sa)
socket.timeout: timed out

from machine-learning.

Our problems need to be further investigated based on the following:

root@ubuntu-xenial:/vagrant# nmap -sS -p8585 xx.xx.xx.xx

Starting Nmap 7.01 ( https://nmap.org ) at 2018-09-29 22:01 UTC
Nmap scan report for xx.xx.xx.xx (xx.xx.xx.xx)
Host is up (0.0020s latency).
PORT     STATE SERVICE
8585/tcp open  unknown

Nmap done: 1 IP address (1 host up) scanned in 0.67 seconds
root@ubuntu-xenial:/vagrant#
root@ubuntu-xenial:/vagrant#
root@ubuntu-xenial:/vagrant#
root@ubuntu-xenial:/vagrant# nmap -sS -p9595 xx.xx.xx.xx

Starting Nmap 7.01 ( https://nmap.org ) at 2018-09-29 22:01 UTC
Nmap scan report for xx.xx.xx.xx (xx.xx.xx.xx)
Host is up (0.00072s latency).
PORT     STATE    SERVICE
9595/tcp filtered pds

Nmap done: 1 IP address (1 host up) scanned in 0.77 seconds

from machine-learning.

We sent a POST request from the nginx-api, to the webserver-api container:

>>> import requests
>>> requests.post('http://webserver-api:6001/login', headers={'Content-Type': 'application/json'}, data={'user[login]': 'jeff1evesque', 'user[password]': 'password123'}, verify=False)
<Response [400]>
>>> requests.post('http://webserver-api:6001/login', headers={'Content-Type': 'application/json'}, data={'user[login]': 'jeff1evesque', 'user[password]': 'password123'})
<Response [400]>

from machine-learning.

Posting from the nginx-api, to the webserver-api container, without headers:

>>> requests.post('http://webserver-api:6001/login', data={'user[login]': 'jeff1evesque', 'user[password]': 'password123'})
<Response [500]>

from machine-learning.

Even after executing the above, no logs entries exists:

root@webserver-api:/var/log/webserver# ls -l
total 0
-rw-r--r-- 1 root root 0 Jun 21 05:54 flask.log
root@webserver-api:/var/log/webserver# cd /var/machine-learning/log
root@webserver-api:/var/machine-learning/log# ls -l
total 12
-rw-r--r-- 1 root root  117 Sep 12 20:54 __init__.py
-rw-r--r-- 1 root root 5670 Sep 12 20:54 logger.py

from machine-learning.

We'll need to fix, and ensure that some logging mechanism is capable of capturing errors associated with our earlier 500 errors. However, before proceeding, we'll likely need to fix #3287. This requirement will be dictated by the minimum requirement of matching the puppet-agent with xenial, rather than trusty. Then, additional updates associated with python3 + pip3 will be non-required benefits. However, if python3 is implemented, we'll likely need to refactor python code. For example, iteritems will need to be refactored to items.

from machine-learning.