I'm struggling to think of a model where it makes sense to update a primary key as a typical operation.  It suggests, as Adil said, that you may be reasoning wrong about your data model.  Maybe you can explain your problem in more detail - what kind of thing has you updating your PK on a regular basis?

Logged batches do get you pretty close to being able to accomplish what you need, but as you pointed out, they are a higher cost to your cluster (they involve a Paxos round, and generally this is about 3x the cost as a typical update).

On Fri, Feb 6, 2015 at 11:19 PM, Ajaya Agrawal <ajku.agr@gmail.com> wrote:
On Fri, Feb 6, 2015 at 7:43 PM, Adil <adil.chabaq@gmail.com> wrote:
it seems you are doing some thing wrong in your model, why can you go with updating columns of key1 instead of deleting/inserting row key?

You can't update a primary key column with update statements.