Introduce retry mechanisms for kafka and etcd

When infra services like kafka/etcd are unavailable for a brief moment(which can happen due to various resons like a netwrok blip or kafka going for a rebalance) , the lib code has to retry for a certain amount of time before giving up. 

The timeout value has to be configurable from the helm chart and also an optional parameter.

Introduce retries in the kafka/etcd lib to retry as per the given values