Tungsten University
TungstenCluster Master Class
Intermediate

Monitoring & Troubleshooting

Learn how to monitor Tungsten Cluster health effectively and troubleshoot issues across the Manager, Replicator, and Connector layers using cctrl, trepctl, packaged Nagios-style scripts, logs, and performance metrics to resolve incidents quickly.​

This session explains datasource and replicator states, shows how to read cctrl status for roles/latency, and pinpoints problems using the logs with actionable error patterns.​

You’ll also learn to gather diagnostics for support with tpm diag and tungsten send diag, analyze replication latency causes, and use trepctl perf to localize bottlenecks by stage with refreshable, per-stage timing stats.

Topics Covered

  • 00:00 Introduction
  • 00:14 Topics
  • 00:30 Monitoring Cluster Health
  • 03:02 Datasource States
  • 04:20 Replicator States
  • 05:20 Getting Fancy With cctrl
  • 05:36 check_tungsten_* Scripts & Examples
  • 07:05 Log Files
  • 07:52 Mining for Errors
  • 09:41 Gather Log Files for Support
  • 10:36 Diagnosing Replication Latency
  • 15:30 The API
  • 16:34 Summary
  • 17:25 Thank you