Fix(tests): start local agent by default

[dracher]

update all test cases(which requires 2 agents) to use node-local as second agent remove redundant agent configs and use the controller machine for local agent operations simplify test setups by removing unnecessary agent configuration

current progress:

• [x] monitor-wildcard-unit-changes
• [x] proxy-service-start
• [x] bluechi-reset-failed
• [x] monitor-multiple-nodes-and-units
• [x] monitor-system-status 3 agents
• [x] bluechi-nodes-statuses
• [x] bluechi-is-oneline-system
• [x] bluechi-reset-failed
• [x] bluechi-reset-failed-node
• [x] proxy-service-stop-target
• [x] proxy-service-stop-requesting
• [x] proxy-service-stop-bluechi-dep
• [x] bluechi-list-units-on-all-nodes
• [x] monitor-wildcard-node-reconnect
• [x] bluechi-is-online-system-monitor
• [x] monitor-multiple-nodes-and-units
• [x] proxy-service-fails-on-execstart
• [x] proxy-service-fails-on-typo-in-file
• [ ] ~bluechi-agent-invalid-port-configuration~
• [x] proxy-service-multiple-services-one-node
• [x] service-fails-on-non-existent-service
• [x] service-stop-requesting-with-unneeded
• [x] multiple-services-multiple-nodes
• [x] propagate-target-service-failure
• [x] socket-activation-parallel

Fixes: #1042

[coveralls]

coverage: 82.331% (+0.07%) from 82.258% when pulling 0e36bba0b27c95467f9366816bca0433b0a0fabc on dracher:VROOM-26373 into 55d191ae0f70f8950fc5fd3a06d2e60bf4e4a580 on eclipse-bluechi:main.

[engelmi]

@dracher Any updates on this?

[dracher]

Hi @engelmi , could you please take a look at the failing error?

[engelmi]

/packit test

[engelmi]

Hi @engelmi , could you please take a look at the failing error?

@dracher Have you run the tests locally? I am not sure what broke, the error with podmans connection pool for the UDS is quiet weird. Retriggered the packit tests, just to verify these still work. If yes, then its some infra issue on GH side.

[dracher]

@engelmi yes i tested locally, and also all tests passed on testing-farm without errors

[engelmi]

@engelmi yes i tested locally, and also all tests passed on testing-farm without errors

Yes, logs in the newly triggered TF runs look good as well. Building the integration-test-base image and lets see then. I remember the integration tests on GH have passed previously as well, lets see.

Edit: Aaaand quay is having issues so pushing new images fails at the moment.