IBM Support

CCID behaviour with DB2 11 extended RBA

Question & Answer


Question

How are CCID's handled by CDC V10.2.1 and DB2 V11 RBA's as opposed to LRS?

Answer

Prior to DB2 V11, RBAs are 6 bytes long.

Both &CCID and a BIGINT are 8 bytes. All possible values of RBA should fit in &CCID and a BIGINT
without becoming negative, unless some sort of manipulation is being used to extend the high order bit (the sign bit) for the additional two bytes to make the BIGINT value negative.

With CDCzOS V10R2M1, the decision was taken by the Architecture team to not extend the &CCID and &SEQNO journal control fields from 8 bytes to 10 bytes.

The justification of this decision was that the actual value of the field was not so important to a customer as the fact that it had changed from one value to another. As RBA values increased past the 8 byte capability of the journal control field, the lower 8 bytes of the value would be provided in the journal control field, so that when it changed, that change would be detectable.

[{"Product":{"code":"SSTVMA","label":"InfoSphere Data Replication for DB2 for z\/OS"},"Business Unit":{"code":"BU059","label":"IBM Software w\/o TPS"},"Component":"--","Platform":[{"code":"PF035","label":"z\/OS"}],"Version":"10.2.1","Edition":"All Editions","Line of Business":{"code":"LOB10","label":"Data and AI"}}]

Document Information

Modified date:
16 June 2018

UID

swg21986066