This document discusses near real-time data synchronization between a primary datastore like MongoDB and Elasticsearch by tailing the primary's operation logs. It explains why this is useful for faster search, analytics and aggregations. Key terms are introduced like replica sets, oplogs and tailable cursors. Concepts covered include locating the logs, reading them continuously and putting the data into scalable stores. Considerations for implementation involve preventing overload, failover and minimizing lag. A demo of oplog tailing is provided.