You have an average key-value pair size of 100 bytes. Your primary access is random needs on
the table. Which of the following actions will speed up random reading performance on your
Turn off WAL on puts
Increase the number of versions kept
Decrease the block size
Increase the block size
Larger block size is preferred if files are primarily for sequential access.
Smaller blocks are good for random access, but require more memory to hold the block index, and
may be slower to create
Reference: Could I improve HBase performance by reducing the hdfs block size?