Tungsten University
TungstenCluster Master Class
Advanced

Replicating Out of a Cluster

Learn how to replicate out of a Tungsten Cluster using a Cluster-Extractor, which consumes THL from the cluster and applies it to a standalone target for backups, reporting, or heterogeneous integrations without adding that target to the cluster.​

This session explains architecture and use cases, supported targets (MySQL, Oracle, PostgreSQL, Redshift, Vertica, Hadoop, Kafka, MongoDB, Aurora, and more), and how the standalone Replicator remains cluster-aware while enabling filtering and bi-directional options for MySQL↔MySQL.​

You’ll review a working configuration: cluster-alias service naming, host prerequisites and /etc/hosts, relay/cluster-slave topology, replication-host for remote apply, exporting the tracking schema, and safe provisioning from a SHUNNED node with the Replicator offline.

Topics Covered

  • 00:00 Introduction
  • 00:20 Topics
  • 00:32 What is a Cluster-Extractor?
  • 02:09 Why use a Cluster-Extractor?
  • 02:47 Possible Targets
  • 03:16 Config Review
  • 05:41 Demo
  • 15:25 Summary
  • 15:39 Thank you