mistake PR#451
Closed
zpetrace wants to merge 48 commits into
Closed
Conversation
* Card ID: CCT-238 Since BASIC authentication has reached its EOL in insights services, it is no longer relevant or supported in the client tool. This update removes all references to BASIC authentication from the client configuration and documentation.
This branch now only targets 'el8' and 'el9', it is not necessary to run on newer systems, as they start to diverge.
This branch currently targets only EL 9. Signed-off-by: Pino Toscano <ptoscano@redhat.com>
This branch targets only EL 9. Signed-off-by: Pino Toscano <ptoscano@redhat.com>
The key may be missing completely from the response.
* Card ID: CCT-891 * Card ID: RHINENG-13984 insights-client that's not registered should not be able to check in. Technically, it is possible to do that. When subscription-manager is registered and the host is pushed to Inventory, its UUID is used for matching. insights-client will collect all canonical facts, including its own and sub-man's UUID, and send them in the checkin request. Inventory is able to match sub-man UUID, and doesn't care where the request came from. The end-user experience was confusing, so in Core 3.4.25+, we ensure insights-client itself is registered. See RedHatInsights/insights-core@00e3d5f.
The pytest fixture varies both the name of the temporary directory and the parents of it. By only handling the name of the directory, we had a conflict when we run the test multiple times in a row.
* Card ID: CCT-1059 Status code returned by `insights-client --status` indicates whether a host is registered or not. This change updates the failing integration tests.
Some tests may fail on setup phase, when e.g. subscription-manager fails to register. Setting log level to debug will allow us to see logs from pytest-client-tools, and better investigate the cause for the failure. Cherry-picked from 8f82241.
Cherry-picked from 287c92f.
* Card ID: CCT-1198 * Card ID: RHINENG-7581 The format of collected archive is being cleaned up, with most of the files directly in the archive root being removed or moved. Files tags.json and branch_info are present twice: once at / and once at /data/. Cherry-picked from 07e0353.
By replacing --output-file with --output-dir, we can omit manipulations with the tar file. Cherry-picked from 1ac9bdb.
Import the 'constants' module earlier in a global context: this way more things can be added to it later. There should be no behaviour changes. Signed-off-by: Pino Toscano <ptoscano@redhat.com> (cherry picked from commit 685e03a)
* Card ID: CCT-226 Since the argument for the --diagnosis parameter was removed from insights-core code, it is no longer relevant to keep the reference to it in the documentation.
(cherry picked from commit 034af90)
* Card ID: CCT-1199 When HTTP_PROXY environment variable is set, and is pointing at valid proxy server, these two tests would fail because insights-client reports different errors on network errors for proxy/non-proxy scenarios. This code takes proxies into account, and asserts strings accordingly. (cherry picked from commit 4996580)
* Card ID: CCT-1220 On systems where insights-client hasn't been run yet (e.g. pure container which only has the repository and test dependencies installed), the test fails because it never gets to the mocked section. File existence and GPG validation in 'run_phase()' are only done if 'config.gpg' is set to True; mocking the attribute to read False skips these checks. (cherry picked from commit 6f6337c)
* Card ID: CCT-1195 When the host is not registered to insights-client, the output for diagnosis argument should not print out 404 error message. This change updates the failing integration test.
* Card ID: CCT-1122 Previously, certain commands (e.g., `--status`) could unintentionally trigger system unregistration if the host was missing from the Inventory. This change adds the integration test which verifies that system unregistration does not happen implicitly.
Some of the test run failed on the timeout, so I created this PR extending the duration of the test runs to 3 hours instead of 2 to prevent failures on timeout.
Adding the traceability docstrings and pytest tier marks into the playbook-verifier test.
As part of the work done for CCT-1237 I have added the debugging options for this test to be able to catch error more efficiently in case of this test failing. We will now be able to also see why the test was actually failing and not just the error from the spec JSON. As a minor change I have also tried to reduce the http_timeout to make this test functional.
As part of a CCT-1237 I have edited 3 different tests that are connected to the archive creation / upload. I have removed the parts that are connected to inventory more than the client side and therefore we won't have to edit the tests every time core or inventory teams change something in future. It now tests the client side only - creation of the archive and that the archive is able to be uploaded. I have also added the dependecies for the bz2 compressor as the test was failing due to missing them. I have also changed checking of display_name to fqdn. This makes the test more reliable in CI environments by comparing the system's fully qualified domain name (FQDN) from the inventory instead of display_name, which can differ based on system or setup. (cherry picked from commit 171184f)
insights-core will get soon an own SELinux policy, and thus insights-client must properly run insights-core using its own policy. To do that: - add a new meson option to specify the type of the SELinux policy for insights-core; an empty string disables the switch, i.e. keep the current behaviour (insights-core running in the same context of insights-client) - use the libselinux Python binding in case a SELinux policy for insights-core is specified - detect whether SELinux is enabled, and disable the switch if not - in case a switch has to happen, then use the existing SELinux context, and apply the different type for insights-core Card: CCT-984 Signed-off-by: Pino Toscano <ptoscano@redhat.com> (cherry picked from commit 169eb50)
- Updates Core to 3.5.13
As a final effort for the CCT-1237 I tried fixing the last failing test. I have fixed test_display_name_disable_autoconfig_and_autoupdate based on the changes we have made in RHEL8 where the test worked after these changes.
* Card ID: CCT-265 This patch updates the tests with logging messages that are shown to the user when the host is not registered. (cherry picked from commit 9690c034c57cbd9381b97cd714e3fc0312da6b7b)
Some of the architectures like ppc64le or s390x are running on emulated systems - they are running on x86_64 with e.g. s390x emulated. This slows the execution of the tests which then usually fail on the timeout. Until that is solved and we have other arches available without having to emulate them we need to extend the duration of the test runs. (cherry picked from commit c6fbfc3)
In order for insights-client to reliably check the correct results from Advisor we needed to change the order of action in the test. The permission of the file is changed before registering to insights so that there is enough time for the issue to be reflected in Advisor. (cherry picked from commit 92538eb)
The test_tags and test_common_specs are currently not running on other architectures other than x86_64. This will need to be fixed some more in the future with better assert. For that we'll need systems not emulated but the exact architecture. Until then, as it is being worked on, we'll skip the tests for now. (cherry picked from commit eafc474)
I have added multiple tests for insights-client-ros. I am testing all the basic functionality of ROS - these tests test that we can download the subpackage and activate the pmlogger. Then we test if the latest metrics exists once the subpaclkage is started. Then they test basic functionality of insights while ROS is installed and active and lastly that we can uninstal the subpackage and the pmlogger is inactive again. This is an effor based on the CCT-1402. (cherry picked from commit cfcd228)
* Card ID: CCT-1204 Since data obfuscation is done by Core now, the .exp.sed file is no longer needed to be present in this repository.
…ouble The original docs/file-redaction.yaml.example doesn't work well as it is because of the tab character at the beginning of the line. Signed-off-by: Masatake YAMATO <yamato@redhat.com> (cherry picked from commit 059184e)
As a hopefully last touch to the tests in an effort to upload test runs through betelgeuse to Polarion we need an parametrized field. It was added to the master branch in PR #415 but the changes were not backported to other branches. As it also includes changes that are no longer needed I have created this separate PR and not backport to fix this field.
Some tests were continuously failing when run against Satellite. The test_set_ansible_host_info was according to related CCT card not fixed on Satellite 6.15 and therefore should be skipped in that version. The test_compliance_policies_option was failing due to API communication troubles so in that case it will be also skipped. Based on some communications the test_display_name is not applicable on Satellite and therefore skipping it entirely if it run against it. This is the final part of CCT-1387. (cherry picked from commit 0046638)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
7 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This was created out of push from a wrong terminal window, sorry, will close right away.