If you are reading and making decisions about what to write just remember there are no transactions. You are essentially running at a Read Uncommitted level of transaction isolation, with regard of batch mutations (a mutation for a single row is atomic).
If you can it may be less headache to write without looking first. Aaron On 28 Aug 2010, at 00:57, Daniel Doubleday <daniel.double...@gmx.net> wrote: > Hi people > > I was wondering if anyone already benchmarked such a situation: > > I have: > > day of year (row key) -> SomeId (column key) -> byte[0] > > I need to make sure that I write SomeId, but in around 80% of the cases it > will be already present (so I would essentially replace it with itself). RF > will be 2. > > So should I rather just write all the time (given that cassandra is so fast > on write) or should I read and write only if not present? > > Cheers, > Daniel
