Fix swallowed failures in Reactive SQL Client tests

[yrodiere]

I have no idea why these tests swallow failures, but it doesn't seem necessary to test the CDI bean producer...
I mean we're not even checking that failures occur, so I really don't know why we would trigger those failures.

I'd like to fix this in order to reuse application-default-datasource.properties in other tests for another PR. Currently it's not possible, since that file is incomplete: depending on the client, sometimes it's completely empty, sometimes it's lacking credentials, ...

Hey @tsegismont can you confirm I didn't overlook some important reason to have connections fail and ignore the failures here?

[yrodiere]

Well the tests pass once we fix the configuration and stop ignoring failures, so I guess this is good to merge?

[gsmet]

I might be mistaken but I wonder if it was to have the tests working even when you don't have a Docker environment.

[tsegismont]

@yrodiere @gsmet To be honest, I can't remember why the failures are ignored.

Looking at the history, I think initially it was convenient to have the tests passing even when no database was available. But, later on, the tests have been disabled by default and can be enabled by profile.

So I believe it's fine to stop ignoring failures now.

[tsegismont]

LGTM, thank you @yrodiere

[yrodiere]

Np, thanks for confirming @tsegismont !

@gsmet would you mind approving? Apparently @tsegismont's approval is not enough :)

[yrodiere]

@gsmet would you mind approving? Apparently @tsegismont's approval is not enough :)

Bump :)

[quarkus-bot]

✔️ The latest workflow run for the pull request has completed successfully.

It should be safe to merge provided you have a look at the other checks in the summary.

You can consult the Develocity build scans.

[yrodiere]

Thanks @gastaldi!