When the Kafka producer is used to forward alarms and these are filtered using an SPEL expression, the ktable synchronization (aka datasync) will identify the filtered alarms as mismatches.

These won't actually get sent to the topic, but these will be reported as additions in the log messages.