January 2020 Summaries

The California Consumer Privacy Act (CCPA) came into effect on January 1, 2020, requiring businesses to inform consumers about any personal data being collected and provide an option for deletion. Ad tech companies are among those affected by the law, which is considered a national one due to its widespread impact. The CCPA aims to give consumers greater control over their privacy and establish trust with customers. Companies must disclose what data they collect, allow consumers to request deletion of their personal information, and avoid discrimination when doing so. A detailed roadmap outlining access to consumer information is also required. Despite some challenges in implementing the law, experts believe it will provide a competitive advantage for companies that comply with it.

This blog post discusses how to implement transactional semantics in an Aerospike application, focusing on handling common read-write transactions and resolving "in-doubt" transactions due to partition changes from network or node failures. It explains that while all single record operations have transactional guarantees in Aerospike, transactions do not span multiple record boundaries. The post provides guidance on using the Read-Modify-Write (or Check-and-Set) pattern and handling in-doubt transactions by maintaining a list of transaction IDs within the affected record. It also covers choosing read consistency levels and managing data that requires consistency across all replicas in a distributed cluster.

This article presents a simplified example of near real-time Ad Campaign reporting using Kafka, Aerospike's edge-to-core data pipeline technology, and Apollo GraphQL. The solution consists of three parts: real-time capture of Ad events via Aerospike edge datastore and Kafka messaging (Part 1), aggregation and reduction of Ad events into actionable Ad Campaign Key Performance Indicators (KPIs) using Aerospike Complex Data Type operations (Part 2), and displaying those KPIs in an Ad Campaign user interface using GraphQL to retrieve data stored in an Aerospike Cluster (Part 3). The companion code is available on GitHub.