These questions were raised many times in this ML and in other sources (blogs, etc.). You can find them with a little effort.
Alex Baranau ------ Sematext :: http://blog.sematext.com/ :: Hadoop - HBase - ElasticSearch - Solr On Wed, Aug 1, 2012 at 1:33 AM, Mohammad Tariq <[email protected]> wrote: > Hello Mohit, > > It is advised not to go above two or three column families. > This is because, when we have many column families, the flushing and > compaction can cause too many needless i/o operations, thus reducing > the performance. > > As far as versions are concerned, default value is 3. Although > we can change it as per our need, but it is not recommended to set the > number of max versions to an exceedingly high level because this will > increase size of the StoreFile greatly. > > Regards, > Mohammad Tariq > > > On Wed, Aug 1, 2012 at 10:10 AM, Mohit Anchlia <[email protected]> > wrote: > > Are there any disadvantages of: > > > > 1. using multiple column families in a table > > 2. using multiple versions of a columns > -- Alex Baranau ------ Sematext :: http://blog.sematext.com/ :: Hadoop - HBase - ElasticSearch - Solr
