Hi, This benchmark study might help you: https://hortonworks.com/blog/hbase-cassandra-benchmark/ However, YMMV as per your workload, cluster tuning and node types. If you are doing random point lookups(using exact rowkey/primary key) then i think you can probably try using https://phoenix.apache.org/language/index.html#small_hint
~Anil Gupta On Thu, Apr 12, 2018 at 2:29 PM, Sawant, Chandramohan < chandramohan.saw...@citi.com.invalid> wrote: > Hi All, > > Is there any optimized way to retrieve the larger dataset of ~1 million > rows (1 row = 200 bytes of data) from HBase via Phoenix jdbc driver. > In my application we are using Phoenix 4.8 with HBase 1.2 and getting > slower performance while fetching the data through sql query. > We have a 5 node cluster with 5 region servers respectively. > > Do let us know if you have encounter this situation and what is the > suggestion we should implement it to make it faster. > > > Regards, > CM > +1 201 763 1656 > > -- Thanks & Regards, Anil Gupta