object-serialization

Objects in Java live in RAM with memory addresses. These addresses only make sense within the same JVM. When we need to send an object over a network (for RMI calls) or store it in a file, we cannot send raw memory — we need a machine-independent representation.

Object Serialization is the process of converting a Java object into a byte stream so it can be transmitted over a network or stored in a file, then reconstructed later. The reverse process (deserialization) reconstructs the object from the byte stream.

1. Class implements Serializable interface
2. During serialization:
   • JVM captures class metadata (class name, serialVersionUID)
   • Recursively serializes all non-static, non-transient instance fields
   • Handles object graphs and circular references
   • Outputs structured byte stream
3. During deserialization:
   • JVM reads byte stream
   • Allocates new memory for object
   • Sets field values from stream
   • Does NOT call constructor
4. For RMI: serialized objects are sent as parameters/returns

• Portable: Platform-independent byte representation
• Recursive: Serializes entire object graph
• Reference tracking: Handles circular references to avoid infinite loops
• Identity not preserved: Deserialized objects are new instances, different from original
• State transfer only: Object data transfers, not memory identity
• serialVersionUID: Version tracking for compatibility

• Built from: Java Standard Edition — Java feature
• Builds into: RMI Remote Method Invocation — RMI depends on serialization
• Builds into: EJB Object — EJB uses serialization
• Related: Transient Keyword — marks fields to exclude
• Contrasts with: JSON Serialization — text-based alternative

• static and transient fields are NOT serialized
• Non-serializable fields cause NotSerializableException
• Constructor is NOT called during deserialization — may skip important initialization
• Identity is lost: original != deserialized (different memory addresses)
• Security risk: Deserialization can execute malicious code
• Performance cost: Slow and CPU/memory intensive