Add In memory database support.

[kartik7153]

As we know Kafka stream consumers are bounded by Kafka partitions and making an I/O call would result in lag if Kafka throughput is very high in that case in memory database support would save the I/O time.