action #44708
closed[sle][functional][y] test fails in enable_usb_repo - Unschedule enable_usb_repo
0%
Description
Observation¶
openQA test in scenario opensuse-Tumbleweed-GNOME-Live-x86_64-gnome-live@USBboot_64 fails in
enable_usb_repo
It seems that USBOOT setting is used in live images tests
https://github.com/os-autoinst/os-autoinst-distri-opensuse/blob/d19a4cee0000c76aa9aa4e878ad20b0d89fcc7cd/lib/main_common.pm#L1082
We need to unschedule this test in some way and it might not be some simple as remove USBBOOT variable of the testsuite as it could be used for other steps in the test suite.
Reproducible¶
Fails since (at least) Build 20181203
Expected result¶
Last good: 20181130 (or more recent)
Further details¶
Always latest result in this scenario: latest
Updated by okurz over 5 years ago
- Due date set to 2018-12-18
- Priority changed from Normal to High
- Target version set to Milestone 21
We should clarify why this starts to fail now.
Updated by riafarov over 5 years ago
- Status changed from New to Workable
- Priority changed from High to Urgent
I somehow expected that we will use that INSTALLATION_VALIDATION variable to enable it, but yeah, let's fix the regressions we have introduced! ;)
Updated by okurz over 5 years ago
https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/6324 was introducing that.
Updated by riafarov over 5 years ago
Fallback from this PR didn't work: https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/6324/files
Updated by okurz over 5 years ago
- Related to action #42917: [sle][functional][y] Introduce only relevant tests modules for USBinstall test suite added
Updated by okurz over 5 years ago
- Status changed from Workable to Rejected
- Assignee set to okurz
let's reopen #42917 as it is a direct regression and the other ticket has more information
Updated by JERiveraMoya over 5 years ago
- Status changed from Rejected to In Progress
Updated by JERiveraMoya over 5 years ago
- Status changed from In Progress to Workable
Updated by JERiveraMoya over 5 years ago
- Assignee deleted (
okurz) - Priority changed from Urgent to Normal
Updated by riafarov over 5 years ago
- Priority changed from Normal to Urgent
- Estimated time set to 3.00 h
EXCLUDE_MODULES variable can be set to reduce priority.
Updated by okurz over 5 years ago
- Priority changed from Urgent to Normal
Other hints: Just evaluate the schedule locally with e.g. isotovideo -d _exit_after_schedule=1 or run clones on production, e.g. clone from o3 to o3.
Updated by favogt over 5 years ago
Ping - this still impacts Leap 15.0, 15.1 and TW.
Why was the regression introducing commit not simply reverted until a proper one is available?
Updated by JERiveraMoya over 5 years ago
It is a regression in the sense that make red the test, but in fact it is not. This test does not make sense to schedule it here, it was doing nothing (enabling a repo that was already enabled and that was not even usb), so therefore we have schedule to work on this test during this sprint as we don't have easy way to unschedule tests (due to main.pm problem we all are aware) without minimal testing/cloning. Besides that the original PR is correct and fix it for sle15sp1 in a test suite where the test module should be scheduled.
Updated by JERiveraMoya over 5 years ago
- Status changed from Workable to In Progress
- Assignee deleted (
riafarov)
Removing USBBOOT
Using EXCLUDE_MODULES=console-enable_usb_repo
note: there is a few console test at the beginning that are not scheduled, I don't know why, perhaps something in my setup when I clone in local but I guess it should not affect.
Updated by riafarov over 5 years ago
- Status changed from In Progress to Workable
- Assignee set to riafarov
EXCLUDE_MODULES can be used to remove failure from the runs on prod, but that's not a proper solution. Here is PR I propose: https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/6375 which fixes original issue we got exposed now.
Updated by JERiveraMoya over 5 years ago
Updated by riafarov over 5 years ago
- Status changed from Workable to Feedback
https://openqa.opensuse.org/tests/812945#
I will keep it open to see if we have missed anything.
Updated by okurz over 5 years ago
JERiveraMoya wrote:
[…] as we don't have easy way to unschedule tests (due to main.pm problem we all are aware)
I think using EXCLUDE_MODULES would be an easy way just as well as excluding the single test module within the main.pm file is. Let's use this as an example for next time to know that actually people care about the test results more than than we might think ;)
Updated by JERiveraMoya over 5 years ago
I was not aware of this feature before. Potential issue that I see is (1) that prevent us to create better solution analyzing other tests suites affected and (2) is easy to "hide" your test failing as result of your PR and there is not trace of that (a progress ticket to discuss about unscheduling it). The reviewer can make a bad decision and makes it more difficult for following reviews as for example remove a test could potentially also affect following tests in a testsuite so comparing test suite is even more tricky. I guess is risky to use it but if properly discuss in the channel during review can help while we find better solution for main.pm.