I would like to convert a column of ARRAY data-type such that each element
of the array is returned as a row. Hive supports it via Lateral Views (
https://cwiki.apache.org/confluence/display/Hive/LanguageManual+LateralView
)?
Does UDF framework in Phoenix allow for building such functions?
You can use local indexes to reduce write amplification. In that case, all
index writes are local writes so the impact of multiple secondary indexes
is not as severe. Of course, there's a read penalty you'd pay, so make sure
you're ok with that.
On Tue, Jan 16, 2018 at 12:08 AM,
For real time writing scenario, the throughput will be down when original
table have more index table. According to the implement of secondary index,
how to accelerate writing speed?