Revise the approach for setting level_compaction_dynamic_level_bytes

[ahamlat]

PR description

We noticed with the upgrade from RocksDB from 8.3.2 to 9.7.3 that the defaults on LeveledCompaction are changed to reduce space amplification. This changes triggered big compactions depending on the nodes when opening RocksDB databse. This is related to change in the value of this parameter level_compaction_dynamic_level_bytes, changed from false to true.
This PR modify the logic for determining level_compaction_dynamic_level_bytes

1. If a node is already sync'd and validating : we keep the existing configuration (level_compaction_dynamic_level_bytes=false if the node is not upgraded yet to 24.12.0, 24.12.1 or 24.12.2) (i.e no compaction)
2. If the node is already ugraded to 24.12.0, 24.12.1 or 24.12.2, we keep the existing configuration, i.e level_compaction_dynamic_level_bytes=true, as the compaction already happened
3. If the node is a new node, sync'ing from scratch, we apply directly the default configuration, i.e : level_compaction_dynamic_level_bytes=true

Tested with success :

1. A node already sync'ed and not upgraded keeps level_compaction_dynamic_level_bytes set to false (no compaction)
2. A node already sync'ed and upgraded keeps level_compaction_dynamic_level_bytes set to true (compaction already happened, so no more overhead)
3. A node sync'ing from scratch, start with the recommended configuration, level_compaction_dynamic_level_bytes set to true

To detect similar issues in the future, we added a warning log if database opening exceeds 60 seconds.

{"@timestamp":"2024-12-16T15:35:11,767","level":"INFO","thread":"main","class":"DatabaseMetadata","message":"Lookup database metadata file in data directory: /data/besu","throwable":""}
{"@timestamp":"2024-12-16T15:35:11,791","level":"INFO","thread":"main","class":"RocksDBKeyValueStorageFactory","message":"Existing database at /data/besu. Metadata versionedStorageFormat=BaseVersionedStorageFormat{format=BONSAI, version=3}. Processing WAL...","throwable":""}
{"@timestamp":"2024-12-16T15:36:11,980","level":"WARN","thread":"pool-2-thread-1","class":"RocksDBOpener","message":"Opening RocksDB database is taking longer than 60 seconds... This may be due to prolonged RocksDB compaction. Please wait until the end of the compaction. No action is needed from the user.","throwable":""}
{"@timestamp":"2024-12-16T15:37:13,740","level":"INFO","thread":"main","class":"FlatDbStrategyProvider","message":"Flat db mode found FULL","throwable":""}

Fixed Issue(s)

Thanks for sending a pull request! Have you done the following?

• Checked out our contribution guidelines?
• Considered documentation and added the doc-change-required label to this PR if updates are required.
• Considered the changelog and included an update if required.
• For database changes (e.g. KeyValueSegmentIdentifier) considered compatibility and performed forwards and backwards compatibility tests

Locally, you can run these tests to catch failures early:

• unit tests: ./gradlew build
• acceptance tests: ./gradlew acceptanceTest
• integration tests: ./gradlew integrationTest
• reference tests: ./gradlew ethereum:referenceTests:referenceTests

[macfarla]

nice. changelog entry?