QACTL review and fixes

[jmv74211]

The objective of this issue is to give a general review of the qactl tool for this first iteration, and to note any possible bugs or improvements needed to meet the main objectives of the tool.

Improvements

• Validation of the input parameters of the qactl configuration (QACTL: Validate configuration file input parameters #1711)
• Automatic download of customized vagrant boxes if necessary (QACTL does not download the boxes needed #1720)
• Improve docker image generation. Right now it is necessary to copy entrypoint and Dockerfile to a custom directory and from there generate the image. (QACTL: Improve Docker image generation process #1749)
• Add healthcheck parameter to wazuh deployment stage. If True health-check will be performed, False otherwise.
• Remove uselss configuration keys: wazuh_installation from wazuh deployment
• Rename ansible tasks in run test module.
• QACTL logging (QACTL logging #1713)
• Install wazuh via S3 packages (QACTL: Improvements for qactl tool implementation #1768)
• Parallelize QACTL modules (Parallelization of QACTL modules #1754)
  • Parallelize deployment and destruction of instances.
  • Parallelize the provisioning of instances.
  • Parallelize the testing
• Clean tmp files (QACTL: Improvements for qactl tool implementation #1768)
• Rename QACTL modules: Names have to be in format snake_case (QACTL: Improvements for qactl tool implementation #1768)
• Update and create new QACTL generic documentation (QACTL : Create a new QACTL documentation #1793)
• Write a documentation for all the modules that aren't currently documented QACTL: Document all QACTL modules #1798

Fixes

• Centos_8 box is not accessible by SSH
• Run test module can not be run independently. QACTL: Separate run test module from the rest of the modules #1734
• Deployment stage QACTL: fix docker deployment and provisioning #1743
  • Containers are not deleted after test execution. It is necessary to kill them.
  • Connection problems with docker instances.
    • Amazon Linux instance inaccessible by ssh.
    • Ansible cannot connect to the Ubuntu instance, although we can connect via SSH.
• Run test stage (QACTL: Fix HTML report generation and update documentation #1728)
  • HTML report is not generated correctly. File not found error.
  • Update documentation. The issue documentation is outdated..
• Fix ansible connection issue (#C4)

Manual checks

• Deployment:
  • SSH host key checking 🟢 (#C2)
  • Vagrant
    • Deploy 1 to n instances using Ubuntu 20.04 custom box 🟢
    • Deploy 1 to n instances using CentOS 8 custom box 🟢
  • Docker
    • Deploy 1 to n instances using Amazon Linux 2 custom image 🟡
    • Deploy 1 to n instances using Ubuntu 20.04 custom image 🟡
• Provision:
  • Provision instance (#C4) Fixed 🟢
  • Provision n Ubuntu 20.04 instances 🟢
  • Provision n CentOS 8 instances 🟢 (#C3) Fixed
  • Provision wazuh via sources 🟢 (#C1) Fixed
  • Provision wazuh via local packages. 🟢
  • Provision qa_framework with custom branches 🟢 .
  • Provision local machine 🟢

[jmv74211]

C1: Provision wazuh via sources

Error

TASK [Render the "preloaded-vars.conf" file] ***********************************
fatal: [172.16.1.60]: FAILED! => {"changed": false, "checksum": "961aba88b0af13a2bd3c5b7e6d2c475f36c116ba", "msg": "Destination directory /tmp/wazuh-v4.2.0-rc13/etc does not exist"}

Fix: da82a81

---

C2: SSH host key checking

Error

fatal: [172.16.1.70]: UNREACHABLE! => {"changed": false, "msg": "Failed to connect to the host via ssh: @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r\n@    WARNING: REMOTE HOST IDENTIFICATION HAS CHANGED!     @\r\n@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r\nIT IS POSSIBLE THAT SOMEONE IS DOING SOMETHING NASTY!\r\nSomeone could be eavesdropping on you right now (man-in-the-middle attack)!\r\nIt is also possible that a host key has just been changed.\r\n

The solution is to replace in the Ansible inventory the SSH option StrictHostKeyChecking=no by UserKnownHostsFile=/dev/null

---

C3: Provision CentOS 8 instance

Error

fatal: [172.16.1.70]: UNREACHABLE! => {"changed": false, "msg": "Failed to connect to the host via ssh: [email protected]: Permission denied (publickey,gssapi-keyex,gssapi-with-mic).", "unreachable": true}

fatal: [172.16.1.70]: FAILED! => {"changed": false, "checksum": "8af5f9d1fec06422f33251e31541cb4226e4ec6d", "msg": "Aborting, target uses selinux but python bindings (libselinux-python) aren't installed!"}

fatal error: Python.h: No such file or directory

Fixs applied to the box, redone and uploaded to S3

---

C4: Ansible SSH connection UNREACHABLE

Error

fatal: [172.16.1.70]: UNREACHABLE! => {"changed": false, "msg": "Failed to connect to the host via ssh: ssh: connect to host 172.16.1.70 port 22: Connection timed out", "unreachable": true}

Fixed in 0a49d21