The European Spallation Source will produce more data than existing neutron facilities, due to higher accelerator power and to the fact that all data will be collected in event mode with no hardware veto. Detector data will be acquired and aggregated with metadata coming from sources such as sample environment, choppers and motion control. To aggregate data we will use Apache Kafka with FlatBuffers serialisation. A common schema repository defines the formats to be used by the data producers and consumers. The main consumers we are prototyping are a file writer for NeXus files and live reduction and visualisation via Mantid. A Jenkins-based setup using virtual machines is being used for integration tests, and physical servers are available in an integration laboratory alongside real hardware. We present the current status of the data acquisition pipeline and results from the testing and integration work going on at the ESS Data Management and Software Centre in collaboration with in-kind and BrightnESS partners.