feat(chaos): Bump duration of experiments

[merkata]

This PR addresses a timeout in the chaos step of the CI/CD pipeline where the discourse charm does not become healthy in the specified duration (default of 60 seconds). A local test confirms the charm becomes stable, so with this change, the total duration of the chaos experiments is increased, so that the test can (hopefully) pass.

With fixing the pipeline run, some additional steps were needed:

• there is now a cleanup of S3 related variables in the integration test for S3
• I had to add dependencies for the linting to pass
• Functions in the unit and integration tests are refactored in order for pylint to pass
• Some pylint issues are disabled, please have a look and supply feedback
• I added a trivy.yaml config fail to ensure trivy doesn't break the pipeline and created an issue to address trivy findings
• I added a workflow action to sync Github issues to our Jira (so that the issue above lands in Jira as well)

NOTE I've cherry picked commits based on scope and have created additional PRs for Jira, Trivy and linting, so once these are merged, this PR will be slimmer and will have only Chaos test related changes. Respective issues (linting, trivy) are fixed in the respective PRs. Thanks!

[arturo-seijas]

LGTM

[arturo-seijas]

LGTM

[arturo-seijas]

LGTM

[arturo-seijas]

LGTM

[gregory-schiano]

LGTM

[gregory-schiano]

I'm not a big fan of leaving so many disabled line-too-long, wasn't there a better approach for these ?

[merkata]

I'm not a big fan of leaving so many disabled line-too-long, wasn't there a better approach for these ?

I did try finding a middle ground between black and pylint, did not succeed. To me it is trial and error - bend a line somewhere, see whether black bends it elsewhere, watch pylint complain again. There are 8 ignores, I'd leave this as an issue in the backlog with a good-first-issue label for someone to tackle. I'm too old for this.

[jdkandersson]

This looks weirdly similar to another PR: #33

[merkata]

This looks weirdly similar to another PR: #33

Yes, they are indeed similar - this PR here is the base one and I split it in 4 different PRs with cherry picks - one about Jira sync, one about Trivy config, one about linter issues and that one here related to the Chaos tests. This would be the last PR to merge and with syncing from main, only the chaos related changes will be here.

[merkata]

I'm not a big fan of leaving so many disabled line-too-long, wasn't there a better approach for these ?

Well @jdkandersson is also not a big fan of line-too-long and left a bunch of great feedback on the feat/linter PR, so if you have a look at that, you'll see there are no more line-too-long disables :) Thanks for commenting!

[github-actions]

Test coverage for 922eb89

Name           Stmts   Miss Branch BrPart  Cover   Missing
----------------------------------------------------------
src/charm.py     180     21     56      4    88%   279->283, 376-377, 404-406, 428-432, 441-452, 463->exit, 478-480
----------------------------------------------------------
TOTAL            180     21     56      4    88%

Static code analysis report

Run started:2023-02-02 13:41:29.563146

Test results:
  No issues identified.

Code scanned:
  Total lines of code: 1196
  Total lines skipped (#nosec): 4
  Total potential issues skipped due to specifically being disabled (e.g., #nosec BXXX): 0

Run metrics:
  Total issues (by severity):
  	Undefined: 0
  	Low: 0
  	Medium: 0
  	High: 0
  Total issues (by confidence):
  	Undefined: 0
  	Low: 0
  	Medium: 0
  	High: 0
Files skipped (0):

[gtrkiller]

LGTM

[amandahla]

LGTM

[jdkandersson]

58 comments and 46 commits later 😉