HBase was designed to be simple. It was not the goal of HBase to replace a traditional RDBMS. Therefore, HBase does not support the concept of Secondary Indexes, advanced Query capabilities or even transactions. Secondly, HBase will perform if you have huge datasets. If you just have thousands or even millions of rows, it is better to consider a traditional Database as it will give you more flexibility with working through the data.

The main goal of HBase was to provide low-latency reads and writes to your data that is stored in a columnar fashion. HBase stores its data internally on HDFS. The main difference between HBase and HDFS is that HDFS is not built for fast record lookups in files. That's the gap that HBase fills. HBase creates indexed StoreFiles on HDFS that enables it to perform fast lookups by row keys.