-
Epic
-
Resolution: Done
-
Normal
-
None
-
None
-
None
-
None
-
[Tech Debt] Telemeter integration test
-
False
-
None
-
False
-
Not Selected
-
NEW
-
To Do
-
NEW
-
0% To Do, 0% In Progress, 100% Done
-
MON Sprint 250, MON Sprint 251, MON Sprint 252
Epic Goal
- Improve telemeter flaky integration test
Why is this important?
Ci is unstable and root issues are hard to debug this leads to time consumption task. Some binaries are old and the structure of mainly integration test are hard to understand and investigate issues.
Iissues after rhel update:
Also update binaries and bump go versions is usually hard. When develop new functions or add bugfix sometimes CI rise errors:
https://github.com/openshift/telemeter/pull/526
https://github.com/openshift/telemeter/pull/510
Scenarios
https://github.com/openshift/telemeter/blob/master/test/integration.sh
https://github.com/openshift/telemeter/blob/master/test/integration-v2.sh
Open questions:
- Integration test are needed in order to check robustness of code?
- Are v1 and v2 needed, are testing same thing?
- Should we refactor that tests?
- Other solutions?
1.
|
Update images |
|
Closed | 
|
Mario Fernandez Herrero |
2.
|
Investigate new ci issue |
|
Closed |
|
Mario Fernandez Herrero |
3.
|
Issues bump golang |
|
Closed |
|
Mario Fernandez Herrero |
4.
|
Issues updating tools and go tools |
|
Closed |
|
Mario Fernandez Herrero |
5.
|
Refactor integration test |
|
Closed |
|
Philip Gough |