On Thu, Feb 5, 2015 at 4:15 PM, Andreas Finke <andreas.fi...@solvians.com> wrote:
> we are currently writing the same column within a row multiple times (up > to 10 times a second). I am familiar with the concept of tombstones in > SSTables. My question is: I assume that in our case in most cases when a > column gets overwritten it still resides in the memtable. So I assume for > that particular case no tombstone is set but the column is replaced in > memory and then the 'newest' version is flushed to disk. > Memtables are periodically flushed; some percentage of these "bursts" will always cross a flush boundary, leading to at least two writes to disk for the same exact value. This is without considering any concerns of the class that Johnathan Haddad mentions. I personally would be conceptually uncomfortable with such a design as having a strong smell of Doing It Wrong, but you should be able to design a test which illustrates how badly (or not?) it bloats your SSTables with garbage in actual operation? Or Is writing the same column an an anti-pattern? > Writing the same column 10 times in a second is likely to be an anti-pattern for a log structured data-store with immutable data files. Consider a memory oriented database? =Rob
