199 Commits

Author SHA1 Message Date
Sergio Cazzolato
d532c31da1 Moving to manual opensuse 15.2
As openSUSE Leap 15.2 reach its end of life on Dec. 31, 2021
2021-12-15 14:24:20 -03:00
Sergio Cazzolato
721bed5783 Run spread tests in fedora-35 2021-11-29 16:14:25 -03:00
Ian Johnson
d8d79c823a .github/workflows/test.yaml: restore failing of spread tests on errors (nested)
We need to use pipefail, otherwise the piping to tee will mean that the exit
code is always 0 and thus never actually fails the GitHub Action even when the
spread tests fail.

Signed-off-by: Ian Johnson <ian.johnson@canonical.com>
2021-11-29 05:18:21 -06:00
Sergio Cazzolato
7d688c455d Fix the filter used in jq to send debug info 2021-11-17 16:19:36 -03:00
Sergio Cazzolato
218defb9a4 Send debug info for tests as well 2021-11-17 10:51:28 -03:00
Sergio Cazzolato
bc47fe67f8 Merge branch 'master' into tests-report-spread-errors 2021-11-11 09:35:14 -03:00
Sergio Cazzolato
5e2994123a Merge remote-tracking branch 'upstream/master' into tests-report-spread-errors 2021-11-10 08:43:04 -03:00
Maciej Borzecki
bb32cba643 github: leave a comment documenting reasons for pipefail
Thanks to @mardy for the suggestion.

Signed-off-by: Maciej Borzecki <maciej.zenon.borzecki@canonical.com>
2021-11-10 12:37:57 +01:00
Maciej Borzecki
ab49bd7f6a github: move pipefail and spread into a subshell
Signed-off-by: Maciej Borzecki <maciej.zenon.borzecki@canonical.com>
2021-11-10 07:51:32 +01:00
Maciej Borzecki
6f9aec31a6 github: enable pipefail when running spread
It seems that some recent github runs did not fail although no tests were run.
The only reasonable explanation being either broken action, or missing pipefail
that hid spread's non-0 exit status.

Signed-off-by: Maciej Borzecki <maciej.zenon.borzecki@canonical.com>
2021-11-09 15:53:54 +01:00
Sergio Cazzolato
17986c5b2b Avoid sending empty lines to mongo db 2021-11-05 16:37:30 -03:00
Sergio Cazzolato
76fa9be513 Update how the log parser is used 2021-11-05 10:41:22 -03:00
Sergio Cazzolato
4f5eed80b6 Fix spread command to create spread log 2021-11-04 23:42:26 -03:00
Sergio Cazzolato
6caf7148a3 New workflow step used to report spread error to mongodb
The change includes:
 . New tool to report errors to mongodb
. New workflow step used to collect spread errors and invoque the new
tool
2021-11-04 18:37:27 -03:00
Sergio Cazzolato
2242f5471b Save the spread log to be used by the log parser 2021-11-03 12:18:00 -03:00
Sergio Cazzolato
861bec1619 Adding back debian 10 2021-10-19 13:55:38 -03:00
Sergio Cazzolato
8a840eeab6 Run spread tests on debian 11
Also debian 10 is set as manual
2021-10-19 13:44:12 -03:00
Alberto Mardegan
1966ac1083 CI: merge coverage results from different runs 2021-10-05 15:42:40 +03:00
Michael Vogt
833d7f40fd Merge pull request #10828 from sergiocazzolato/tests-cleanup-selfhosted-workspace
tests: cleanup the job workspace as first step of the actions workflow
2021-09-23 16:50:28 +02:00
Sergio Cazzolato
d8bd3f3753 Remove ubuntu-21.10-64-cgroupv2 from tests workflow 2021-09-22 18:35:50 -03:00
Sergio Cazzolato
a2aa553ebd Cleanup the job workspace as first step
The idea is to start with a clean workspace in the self hosted runner

The previous job could left the workspace in a bad state as it is
happening when it adds a submodule.
2021-09-22 13:50:36 -03:00
Michael Vogt
9e0a1807b3 Merge pull request #10803 from bboozzoo/bboozzoo/cgroupv2-test-and-systems
tests, interfaces/builtin: introduce 21.10 cgroupv2 variant, tweak tests for cgroupv2, update builtin interfaces
2021-09-22 16:46:22 +02:00
Michael Vogt
9804780799 Merge pull request #10740 from bboozzoo/bboozzoo/fault-injection
osutil: helper for injecting run time faults in snapd
2021-09-21 20:07:11 +02:00
Maciej Borzecki
885366fef1 Merge remote-tracking branch 'upstream/master' into bboozzoo/cgroupv2-test-and-systems 2021-09-21 10:24:21 +02:00
Ian Johnson
4c00f39671 Revert ".github/workflows/test.yaml: bump action-build to 1.0.9"
This reverts commit 5296059c30.

Now that the v1 tag has been updated to point to 1.0.9, we can go back to 
tracking v1 instead of tracking specifically v1.0.9, which will prevent us from
getting updates in the future if there is i.e. a 1.0.10 or something.

Signed-off-by: Ian Johnson <ian.johnson@canonical.com>
2021-09-20 08:53:12 -05:00