Tiny tweaks based on QE feedback for Datasource guide

[MichalMaler]

Based on the QE feedback for the Datasource guide, I am proposing tiny tweaks in the guide and its relevant Javadocs. These need to be backported to 3.2.

The QE request was to add default values or specify if the feature is disabled by default.
Quoting : Properties quarkus-agroal_quarkus.datasource."datasource-name".jdbc.leak-detection-interval and quarkus.datasource."datasource-name".jdbc.max-lifetime do not have default duration.

I updated the Javadocs with the @ConfigDocDefault annotation on relevant methods after an SME consultation with @yrodiere , who shared with me the Agroal defaults we, in some cases, share or replace in Quarkus.

    Duration leakTimeout = ZERO;       ->      `@ConfigDocDefault("This feature is disabled by default.")`
    Duration validationTimeout = ZERO;      (backgroundValidationInterval)
    Duration reapTimeout = ZERO;               (idleRemovalInterval)
    Duration maxLifetime = ZERO;       ->     `@ConfigDocDefault("This feature is disabled by default.")`

Additionally, I updated some links here and there.

[github-actions]

🙈 The PR is closed and the preview is expired.

[MichalMaler]

Hello @yrodiere @gsmet
Please, do you have an idea what can cause these fails?

It is possible to merge this PR if we all agree that my changes can hurt the code?

[yrodiere]

It is possible to merge this PR if we all agree that my changes can hurt the code?

It does look like an unrelated failure.

Though you just pushed again so... let's wait for that run to pass.

[yrodiere]

FWIW this looks like the cause:

2023-11-14T12:54:56.4011232Z Error response from daemon: Head "https://quay.io/v2/quarkus/ubi-quarkus-mandrel-builder-image/manifests/jdk-21": received unexpected HTTP status: 502 Bad Gateway

The build of that module basically crashed, resulting in later steps not being executed, which critically means we never stopped the MariaDB container, triggering cascading failures in following modules since they weren't able to start MariaDB.

@gsmet I wonder if that MariaDB test failure we saw the other day about access rights wasn't a similar (but slightly different) problem: some Maven plugin failed, leading to containers not being stopped, leading to later tests reusing them instead of restarting them (since docker-maven-plugin sometimes does that, I think), leading to the wrong users/roles being present in the database we were testing against.

[quarkus-bot]

Failing Jobs - Building b6654f2

Status	Name	Step	Failures	Logs	Raw logs	Build scan
✖	Native Tests - Spring	Build	Failures	Logs	Raw logs

Failures

⚙️ Native Tests - Spring #

- Failing: integration-tests/spring-data-jpa 

📦 integration-tests/spring-data-jpa

✖ Failed to execute goal io.quarkus:quarkus-maven-plugin:999-SNAPSHOT:build (default) on project quarkus-integration-test-spring-data-jpa: Failed to build quarkus application

[yrodiere]

2023-11-14T15:04:51.8298060Z [ERROR] [error]: Build step io.quarkus.deployment.pkg.steps.NativeImageBuildStep#build threw an exception: java.lang.RuntimeException: Failed to pull builder image 'quay.io/quarkus/ubi-quarkus-mandrel-builder-image:jdk-21'

Definitely unrelated, merging

[rsvoboda]

@MichalMaler please prepare dedicated PR for 3.2 branch. Backport can't be applied cleanly, there are conflits in .java file.

Please backport just chages done in docs/src/main/asciidoc/datasource.adoc, that's probably the safest path for now.

[MichalMaler]

@MichalMaler please prepare dedicated PR for 3.2 branch. Backport can't be applied cleanly, there are conflits in .java file.

Please backport just chages done in docs/src/main/asciidoc/datasource.adoc, that's the safest path for now.

@rsvoboda Sure thing! It will be done tomorrow. Will add you for the review :)