Fix action store -> state store migration tests

[AndersonQ]

What does this PR do?

Fixes the state store migration tests so they actually run and skip them for darwin.
It also changes the migration test to use a golden file when testing the migration from the old action store to the new state store.
Finally, it removes the withFile function as its miss usage caused the store migration tests to not actually run.

Why is it important?

• The tests need to fully run
• Without using a golden file and creating the old action store on the fly, the test would not catch changes in the action model. If the model would to be changed, and it will as part of Scheduled upgrade fails with error parsing version "" if agent restarts before the upgrade starts  #3912, the test would still pass. However the actual migration would not work as expected.

Checklist

• My code follows the style guidelines of this project
• [ ] I have commented my code, particularly in hard-to-understand areas
• [ ] I have made corresponding changes to the documentation
• [ ] I have made corresponding change to the default configuration files
• I have added tests that prove my fix is effective or that my feature works
• [ ] I have added an entry in ./changelog/fragments using the changelog tool
• [ ] I have added an integration test or an E2E test

Related issues

• Relates Scheduled upgrade fails with error parsing version "" if agent restarts before the upgrade starts  #3912

Questions to ask yourself

• How are we going to support this in production?
• How are we going to measure its adoption?
• How are we going to debug this?
• What are the metrics I should take care of?
• ...

[elasticmachine]

Pinging @elastic/elastic-agent (Team:Elastic-Agent)

[elastic-sonarqube]

Quality Gate failed

Failed conditions

0.0% Coverage on New Code (required ≥ 40%)

See analysis details on SonarQube

[pchila]

Code looks reasonable, I am a bit surprised by the 0% new code coverage... are the tests actually running?

Moreover, I think that there's a bit of overlap with a bugfix I merged last year where I was moving unencrypted config to the new encrypted diskstores (as the agent was "forgetting" fleet config because fleet.enc was not found), so maybe we need to clean up a bit. Here's where the bugfix is located

elastic-agent/internal/pkg/agent/cmd/run.go

Lines 194 to 206 in bdefb2c

	// Migrate .yml files if the corresponding .enc does not exist
	// the encrypted config does not exist but the unencrypted file does
	err = migration.MigrateToEncryptedConfig(ctx, l, paths.AgentConfigYmlFile(), paths.AgentConfigFile())
	if err != nil {
	return errors.New(err, "error migrating fleet config")
	}
	// the encrypted state does not exist but the unencrypted file does
	err = migration.MigrateToEncryptedConfig(ctx, l, paths.AgentStateStoreYmlFile(), paths.AgentStateStoreFile())
	if err != nil {
	return errors.New(err, "error migrating agent state")
	}

[AndersonQ]

I am a bit surprised by the 0% new code coverage... are the tests actually running?

the new code is test code, which isn't covered by tests. I don't think it's calculating the overall coverage.
Let me calculate it

[AndersonQ]

Quality Gate failed

Failed conditions

0.0% Coverage on New Code (required ≥ 40%)

See analysis details on SonarQube

As Sonarqube looks for coverage only on NEW code and all the new code is test code, it isn't calculating the overall code coverage increase. Here it's:

• main:

 ~/devel/github.com/elastic/elastic-agent/internal/pkg/agent/storage/store   main  16:22:29                                                                                     1.223s
❯ go test -cover   
PASS
coverage: 61.2% of statements
ok  	github.com/elastic/elastic-agent/internal/pkg/agent/storage/store	0.481s

• bugfix/3912-borken-stat-store (the PR's base branch)

 ~/devel/github.com/elastic/elastic-agent/internal/pkg/agent/storage/store   bugfix/3912-borken-stat-store  16:22:43                                                            0.038s
❯ go test -cover                            
PASS
coverage: 61.2% of statements
ok  	github.com/elastic/elastic-agent/internal/pkg/agent/storage/store	0.476s

• 3912-add-state-store-migration-tests: the PR itself

 ~/devel/github.com/elastic/elastic-agent/internal/pkg/agent/storage/store   3912-add-sta…ration-tests  16:22:53                                                                0.024s
❯ go test -cover                                   
PASS
coverage: 75.2% of statements
ok  	github.com/elastic/elastic-agent/internal/pkg/agent/storage/store	0.523s

@pchila here it is, a 14% increase in code coverage

[AndersonQ]

Moreover, I think that there's a bit of overlap with a bugfix I merged last year where I was moving unencrypted config to the new encrypted diskstores (as the agent was "forgetting" fleet config because fleet.enc was not found), so maybe we need to clean up a bit. Here's where the bugfix is located

elastic-agent/internal/pkg/agent/cmd/run.go

Lines 194 to 206 in bdefb2c

	// Migrate .yml files if the corresponding .enc does not exist
	// the encrypted config does not exist but the unencrypted file does
	err = migration.MigrateToEncryptedConfig(ctx, l, paths.AgentConfigYmlFile(), paths.AgentConfigFile())
	if err != nil {
	return errors.New(err, "error migrating fleet config")
	}
	// the encrypted state does not exist but the unencrypted file does
	err = migration.MigrateToEncryptedConfig(ctx, l, paths.AgentStateStoreYmlFile(), paths.AgentStateStoreFile())
	if err != nil {
	return errors.New(err, "error migrating agent state")
	}

@pchila, good point, it might be messing the store migration. I added a TODO on my currently WIP PR: 4926df0