[fix][ci] Ensure disk headroom for replication tests

[lhotari]

Motivation

Replication tests start multiple local BookKeeper ensembles. With limited free disk space, ledger directories can cross BookKeeper's disk-usage threshold and become unwritable, causing failures unrelated to the replication behavior under test.

Modifications

Run the existing disk-cleanup action for Broker Group 5 as well as resource-tracing runs. The action stops once its existing 20 GiB free-space threshold is met. Test selection, BookKeeper thresholds, Gradle caches and build artifacts are unchanged.

Verifying this change

• Make sure that the change passes the CI checks.

spotlessCheck checkstyleMain checkstyleTest passed. Nine workflow-expression cases verified replication and non-replication groups with resource tracing off/on/full. Actionlint reports no new diagnostics relative to master. Local review found no remaining issues.

Does this pull request potentially affect one of the following parts:

• Dependencies (add or upgrade a dependency)
• The public API
• The schema
• The default values of configurations
• The threading model
• The binary protocol
• The REST endpoints
• The admin CLI options
• The metrics
• Anything that affects deployment

[void-ptr974]

LGTM