[SecuritySolution] [8.11] Fix fleet_integration.ts API integration test flake

[jpdjere]

Fixes: #167056

See comment in ticket.
This is only the fix for 8.11.

Summary

• Changes the installation of the security_detection_engine package from the current Fleet bulk install packages endpoint to use the endpoint that installs a single package, in order to get to a more stable test
• Notice that the errors reported that Kibana machine are socket hangups/timeouts and are solved on retry, so they do not block any pipelines)

For maintainers

• This was checked for breaking API changes and was labeled appropriately

[elasticmachine]

Pinging @elastic/security-detections-response (Team:Detections and Resp)

[elasticmachine]

Pinging @elastic/security-solution (Team: SecuritySolution)

[kibana-ci]

💚 Build Succeeded

• Buildkite Build
• Commit: b2ed3b4

Metrics [docs]

✅ unchanged

To update your PR or re-run it, just comment with:
@elasticmachine merge upstream

cc @jpdjere

[banderror]

@jpdjere Is there any evidence, besides the green CI, that these changes reduce the flakiness?

[banderror]

Nit: I think we should do some refactoring in main in another PR. There are 2 files containing 4 functions for doing something with the fleet package with prebuilt rules.

[jpdjere]

@jpdjere Is there any evidence, besides the green CI, that these changes reduce the flakiness?

No, actually: the issue has been extremely difficult to reproduce, it couldn't even be resurfaced with the flaky test runner. And Kibana Machine was reopening the ticket even when the test passed on a retry. The socket hangups/timeouts errors that were reported pointed to an issue on Fleet serverside (the endpoint timeing out, maybe), so switching to an individual package installation endpoint will hopefully give us a stabler test. There's no firm evidence, but I will keep an eye on the ticket to see if the flake is re-reported on 8.11.

[mistic]

This PR didn't make it into the last BC of 8.11.0. Updating the labels.