Skip to content

Latest commit

 

History

History
98 lines (72 loc) · 3.74 KB

installation-guide-functional-tests-centos7.md

File metadata and controls

98 lines (72 loc) · 3.74 KB
Raw

Installation Guide for the Orion-LD Functional Test Suite on CentOS 7

The functional test suite for Orion-LD, found in the Orion-LD repository, under test/functionalTests, needs some installation in order to work.

First of all, there is a python script (scripts/accumulator-server.py) that is used as the receptor of the notifications during the functional tests.

This python script needs a few packages to be installed:

  1. Install pip As pip is not available in CentOS 7 core repositories, you need to enable the EPEL repository to be able to install pip. The EPEL (Extra Packages for Enterprise Linux) repository provides additional software packages that are not included in the standard RedHat and CentOS repositories.
sudo yum install epel-release
sudo yum install python-pip

Update pip:

pip install -U pip
  1. Install OpenSSL for Python:
pip install pyopenssl
  1. Install Flask:
pip install -U virtualenv
pip install Flask

That should be all for the accumulator.

The test script (test/functionalTest/testHarness.sh) needs to find the accumulator, to be able to start it:

cd <orion-ld repository base directory>
export PATH=$PATH:$PWD/scripts
which accumulator-server.py

The output of the which command should be:

~/git/context.Orion-LD/scripts/accumulator-server.py

Also, the test script uses nc to verify that the broker has started, and bc for simple calculations. Install both of them:

sudo yum install nc
sudo yum install bc

Test it by launching:

test/functionalTest/testHarness.sh

There are over 1250 test cases (each with a number of steps), so, it will take a while. Orion-LD has inherited the functional test suite from orion and added some 250 test cases only for NGSI-LD. If you want to run only the NGSi-LD test cases, run the suite with the -ld option:

test/functionalTest/testHarness.sh -ld

There are lots of command line options for the test suite; use the -u option to see all of them.

In case some functional test case fails, normally due to installation error, and you're not really interested in that case (might be for IPv6 and you aren't interested in IPv6), you can disable test cases.

Functional test cases can be disabled by exporting an environment variable called CB_SKIP_FUNC_TESTS. For example, to disable the test case 'direct_https_notifications.test' (you don't want https notification, so ...), do this:

export CB_SKIP_FUNC_TESTS=0706_direct_https_notifications/direct_https_notifications.test

Note that not only the name of the test case file, but also the directoy where it resides is part of the "identifier". This is so because different functional test case directories can have test case files with the same name.

FYI: after following myself the instructions in the installation guides, the following functional tests failed for me:

  • 0706_direct_https_notifications/direct_https_notifications.test
  • 0706_direct_https_notifications/direct_https_notifications_no_accept_selfsigned.test

Hmmm, something about https seems to be missing in the instructions ...

Looking closer at the errors, it seems clear that it is the accumulator script that is having problems (the accumulator log-file says: AttributeError: 'Context' object has no attribute 'wrap_socket'). I will have to look into this, but for now, I simply do this:

export CB_SKIP_FUNC_TESTS="0706_direct_https_notifications/direct_https_notifications.test 0706_direct_https_notifications/direct_https_notifications_no_accept_selfsigned.test"

Remember to put this environment variable in your startup files (e.g. ~/.bash_profile), so that you don't lose it. Also, the directory added to PATH should be saved in your startup files.