ZMX #78 Optimization

[zstoimchev]

Implementation and enhancements of #78. See issue for more info...

[zstoimchev]

Backlog

Logger

• Add day-rolling — currently the file is opened once at startup with today's but never rolls over if the process runs past midnight
• Add old file pruning — delete log files older than logger.retention.days (property exists in logger.properties but is never read or acted on)
• Add JVM shutdown hook to flush and close FILE_OUT cleanly on exit

Networking

• Fix thundering herd in startPeerMaintenance — currently submits a task for every unknown peer at once; limit to maxConnections - currentConnections
• Track in-progress connections — maintenance ignores peers currently being to, causing duplicate connection attempts across cycles; use a ConcurrentHashMap keyed set to track them
• Increase ServerSocket backlog in Server — default is 50, which gets overwhelmed at higher peer counts; set explicitly to 200+

CLI

• Implement printCircuitRoutes() — method body is empty

Configuration

• Add connection.retry.attempts and connection.timeout.milliseconds to node.peer.test1/2/3.properties — currently missing, falls back to defaults silently
• Implement config value validation in Config — there is a TODO comment for this; at minimum guard all Integer.parseInt calls so a bad property value doesn't crash startup

Cleanup

• Remove logback.xml from src/main/resources — dead file, no SLF4J or Logback dependency in pom.xml, does nothing
• Clean up connectToPeer retry logic — if (retries == attempts) continue is misleading (works accidentally); replace with break or restructure the loop