Hi everyone! 

I'm new to HBase and I had one question about what's the best write
strategy. I have a process that updates a series of rows several times,
and as HBase is modelled after BigTable, I suppose the best course of
action is to group modifications to a row in a single batch, to avoid
row fragmentation (where hbase has to read several row versions to
provide a response. 

Is this correct or I can update rows as the changes come without
worries?

Thanks in advance for your help!



Reply via email to