The following depicts the systems used by TrueSight Intelligence and the data flow and processing.
The architecture diagram represents the data flow and processing in TrueSight Intelligence. All the clusters are scalable and can be increased as per requirement to provide unlimited scalability.
- The data sources send data to TrueSight Intelligence using the available REST APIs.
- The AWS Route 53 acts as a load balancer and forwards the incoming REST requests to the NGINX and NETTY clusters.
- The NETTY cluster push the incoming messages to corresponding queues managed by Kafka. The incoming messages are sorted by data type and tenant.
- The Storm clusters pulls data from the Kafka message queue and processes incoming data in real-time. Storm is distributed and fault-tolerant, and guarantees that every message is processed at least once.
- The SPARK cluster persists the data, check for threshold violations, and computes baselines.
- The Elastic search stores event data and the Cassandra cluster stores metrics data.
- The NGINX and NETTY clusters send the processed data via the AWS Route 53 to the TrueSight Intelligence user interface.