Solvedansible HaProxy drain mode 'bool' object is not callable error

ISSUE TYPE

• Bug Report

COMPONENT NAME

HaProxy Module

ANSIBLE VERSION

ansible 2.4.0.0
  config file = None
  configured module search path = [u'/Users/stewart/.ansible/plugins/modules', u'/usr/share/ansible/plugins/modules']
  ansible python module location = /Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/ansible
  executable location = /Library/Frameworks/Python.framework/Versions/2.7/bin/ansible
  python version = 2.7.13 (v2.7.13:a06454b1afa1, Dec 17 2016, 12:39:47) [GCC 4.2.1 (Apple Inc. build 5666) (dot 3)]

CONFIGURATION

ansible-config dump --only-changed returned nothing. I have changed no config.

OS / ENVIRONMENT

Running from OSX 10.10.5
Trying to control HaProxy 1.5.8 on Debian 8

SUMMARY

I am able to enable and disable HaProxy backends correctly but when I try to use the "drain" state type I get this error. The playbook works when I use enabled and disabled but fails with the below error when I use drain.

It says, 'bool' object is not callable. I might just be configuring this wrong but I believe I am doing everything correctly and using the correct Ansible version.

STEPS TO REPRODUCE

- name: Set server to drain
  hosts: haproxy
  sudo: yes
  tasks:
    - name: Set server1 to disabled
      haproxy:
        state: drain
        host: server1
        socket: /var/run/haproxy.sock
        backend: cluster_www

EXPECTED RESULTS

I expect the HaProxy backend to change to the drain state.

ACTUAL RESULTS

fatal: [10.5.59.15]: FAILED! => {"changed": false, "failed": true, "module_stderr": "Shared connection to 10.5.59.15 closed.\r\n", "module_stdout": "Traceback (most recent call last):\r\n  File \"/tmp/ansible_Xzs48s/ansible_module_haproxy.py\", line 454, in <module>\r\n    main()\r\n  File \"/tmp/ansible_Xzs48s/ansible_module_haproxy.py\", line 450, in main\r\n    ansible_haproxy.act()\r\n  File \"/tmp/ansible_Xzs48s/ansible_module_haproxy.py\", line 410, in act\r\n    self.drain(self.host, self.backend)\r\nTypeError: 'bool' object is not callable\r\n", "msg": "MODULE FAILURE", "rc": 0}

@alikins I looked into the issue today, and it looks like it's caused by a name collision between a variable and function in the module created for handling the new drain state and parameter. I've submitted PR #30947 to hopefully address this, but have not been able to test this against an HAProxy-managed cluster. I probably won't have time to set up/test against an HAProxy cluster (mock or otherwise) for a couple weeks.

The main noticeable change brought by this PR is tweaking the interface so that draining replaces drain as a valid state. (My rationale for this is in the linked PR.)

@stewartwebb: Do you have an HAProxy setup that you wouldn't mind testing the PR changes against? If you change state: drain to state: draining and pull down the PR, it will (hopefully!) work as expected.