Hi Rajiv, Did you ever find out what was causing this issue?
I notice something similar on my Kafka Cluster only that the 95th percentile of the log flush goes above 10 or 20 Min, and then I start to see under replicated partitions. Besides the difference on time, the scenario is quite the same. The infrastrucure is slighlty diffferent as well but i use even bigger machines and the kafka is version 10.0.2.1. Thanks, Martin Cremona.
