Could someone explain what a rowkey is?
I have been able to find definitons of row hash and row id, but not rowkey.

Re: rowkey

rowkey is the combination of partition# plus the rowid. (just like how rowid is the combination of rowhash and uniqueness value)

Re: rowkey

No, rowkey is the rowid minus the uniqueness value. In the absence of a partitioned or value-ordered primary index, the rowkey is simply the row hash.

Re: rowkey

Can anybody pls expalin what a rowkey actually is ?

Is it the combination of Partition Id+Row Hash + Uniqueness Value or
it is just the Row Hash value being given the same name.

Re: rowkey

Joe is right.

rowhash + uniqeness value -> ROWID (pre-V2R5)
partition# + rowhash + uniqeness value -> ROWKEY (V2R5+)

But usually everyone is still using the term ROWID instead of ROWKEY :-)


Re: rowkey

Hi Experts,

as like Row id in the Oracle what is the unique identifier in Teradata apart from primaryindex hash value,

i want to know the  as per this comment ( rowid is the combination of rowhash and uniqueness value) how to fetch  the rowid through Query.



Re: rowkey

HI Everyone,

Could anyone please tell me size of Rowkey size..



Re: rowkey

Depends on how many partitions you have. it you have less then 65535 then your rowkey is going to be 

2 bytes + 4bytes for the Row Hash +  4 bytes for the uniqueness value

More then 65535 partitions and it uses 8 bytes for the internal partition number