[fix][test] Fix flaky ExtensibleLoadManagerTest.startBroker timeout

[lhotari]

Fixes #21875

Motivation

Fix flaky ExtensibleLoadManagerTest.startBroker that times out when restarting brokers after testStopBroker or testIsolationPolicy.

The root cause is that after a broker is stopped and its container is restarted, the ServiceUnitStateTableViewImpl channel fails to start due to a compacted ledger read error ("Invalid unknown tag type: 4", a known issue per PIP-429). The broker process crashes and restarts 2-3 times inside the Docker container before succeeding. The previous 120-second timeout was insufficient for this crash-restart cycle.

Additionally, ignoreExceptions() was hiding all errors, making these failures impossible to debug from test logs.

Modifications

• Replace ignoreExceptions() with explicit try-catch and logging inside the Awaitility condition, so errors are visible in test logs while still allowing retries
• Add informational logging when brokers don't see the expected number of active brokers
• Increase the Awaitility timeout from 120 to 180 seconds to accommodate broker crash-restart cycles
• Minor: early return when pulsarCluster is null, rename shadowed variable

Verifying this change

• Make sure that the change passes the CI checks.

This change is already covered by existing integration tests in the pulsar-loadbalance test group.

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