ARCHITECTED FOR REAL-TIME CLOUD MONITORING

Detect and pinpoint anomalies across infrastructure, applications, and business SLAs.

BATCH VS. STREAMING ARCHITECTURE

Streaming is the only way to get real-time insights in seconds. Other monitoring tools that rely on batch processing take too long to alert on issues. Only a streaming architecture can ingest, analyze, and deliver data quickly enough to keep small issues from impacting your customers in a big way.

Batch is Slow

  • Minutes to fire an alert
  • Unpredictable alert latency
  • Performance degrades at scale
  • Not fast enough for ephemeral architectures

Learn more

Streaming is Real-time

  • Seconds to fire an alert
  • Consistent alert latency
  • Performance at scale, with analytics
  • Instant monitoring of ephemeral architectures

Learn more

Sample Vs. NoSample™️ Architecture

Traditional APM vendors use head-based sampling which makes random decisions at the start of the very first transaction and propagates this decision across service boundaries. Head-based sampling can miss important transactions that you need to troubleshoot your issue. SignalFx’s NoSample architecture captures every transaction, and helps you focus on what matters most to troubleshoot issues before they affect your customers.

Traditional APM Sampling Doesn’t Work for Microservices

  • Head-based sampling misses anomalous trace data
  • Does not provide insights into long-tail p99 issues
  • Root-cause analysis is significantly slower or at worst impossible

Learn More

SignalFx NoSample™️ APM Architecture

  • NoSample tail-based architecture captures every transaction
  • Never miss trace data when troubleshooting long-tail issues
  • Smart Gateway intelligently captures anomalous transactions 
  • Metricize spans and traces for granular analytics 

Learn More