Why two records for the same key???

Question

Hi,
I'm finding that our database contains "duplicate" records even though it is set to disallow duplicated. In particular, from DbDump, here are two keys from our database:
737201037870770c0001000000a6005a001471c678
737201037870770c0001000000a6005a0014711678
You can probably really think of them as
737201037870770c 0001000000a6005a00147 1c678
737201037870770c 0001000000a6005a00147 11678
where only the middle part is "our key data". Use use Externalizable and have methods for the key:
public void readExternal(ObjectInput in) throws IOException, ClassNotFoundException {
          short oldVersion = in.readShort();
          this.primaryKey = in.readInt();
          this.orderPurpose = in.readChar();
          this.foreignKey = in.readInt();
public void writeExternal(ObjectOutput out) throws IOException {
out.writeShort(externalizationVersion);
out.writeInt(this.primaryKey);
out.writeChar(this.orderPurpose);
out.writeInt(this.foreignKey);
public int compare(Object o1, Object o2) {
byte[] b1 = (byte[]) o1;
byte[] b2 = (byte[]) o2;
final int compareLen = b1.length;
for (int i=0; i<compareLen; i++) {
if (b1[i] != b2) {
return (b1[i] < b2[i] ? -1 : 1);
return 0;
My question is basically this: what are the trailing bytes for in the key and why do they differ in my case? BTW, we're using an old version, 2.0.90, and we recently switched to Java 6 from Java 5... (just in the interest of full disclosure!)
Thanks very much in advance,
Dan

Answer 1

Yes, you're right. That hint really helped me find
my problem !!Glad to hear that.
I still don't know what the leading and trailing
bytes are used for but at this point, it doesn't
really matter anymore.They're output by Java serialization when JE calls ObjectOutputStream.writeObject, so they should be defined here:
http://java.sun.com/j2se/1.5.0/docs/guide/serialization/spec/protocol.html
Thanks for your help,You're welcome,
Mark