Even if data is time-ordered there may be locking and contention issues with closely spaced GPS position data. It may be desirable to use Flink in between the Kafka records and the map-matching process within Reporter. Flink could be used to "buffer" single locations into traces.
Even if data is time-ordered there may be locking and contention issues with closely spaced GPS position data. It may be desirable to use Flink in between the Kafka records and the map-matching process within Reporter. Flink could be used to "buffer" single locations into traces.