Skip to main content

PerForge v2.1.5: New Version

· 3 min read
Uladzislau Shklianik
Uladzislau Shklianik
Performance engineer
Siamion Viatoshkin
Siamion Viatoshkin
Performance engineer

PerForge v2.1.5 is a focused quality and insights release on top of the v2.1 line. 🎉

This version sharpens ML/AI analysis, gives you fine‑grained control over how results are interpreted per project, and adds better visibility into transaction status and InfluxDB setups.

1. Smarter ML & AI analysis​

PerForge’s ML engine now does a better job of explaining what went wrong and where.

  • Per‑transaction anomaly analysis ML now looks at metrics per transaction/request and highlights which ones are most likely contributing to a global anomaly. This makes it easier to jump straight to the endpoints that need attention.

  • Red bands instead of red dots Instead of isolated red markers, anomaly periods on graphs are now rendered as red areas. This makes it much easier to see the duration and impact of a problem at a glance.

  • Anomaly windows on per‑transaction graphs For each transaction, PerForge marks the time ranges where anomalies were detected, so you can correlate them with load patterns, deployments, or external events.

  • Anthropic provider support In addition to existing AI providers, PerForge can now use Anthropic (Claude) for AI‑driven summaries and explanations. Configure it the same way as other AI integrations and keep using the same ${ai_summary} / prompt setup.

2. Advanced Settings – per‑project control​

PerForge now exposes the internals of ML and status evaluation through a new Advanced settings page:

  • Navigate to Settings → Advanced settings in the top navigation bar.
  • Settings are stored per project, so each project can have its own ML thresholds, stability criteria, baseline sensitivity, and more.
  • Two main sections are available:
    • ML Analysis Settings – thresholds and parameters for the anomaly detection engine, including per‑transaction coverage.
    • Transaction Status Settings – how NFR checks, baselines, and ML findings are combined into the final Pass / Warning / Failed status.

This is the place to tune PerForge for your specific environment (noisy metrics vs. very strict SLOs, different traffic profiles, etc.).

3. Unified transaction status tables​

We’ve added two new status tables that combine all checks into a single, easy‑to‑read view per transaction:

  • A compact table that shows an overall status by transaction (Pass / Warning / Failed).
  • A detailed table that breaks down status per check type (NFR, baseline comparison, ML anomalies) so you can immediately see why something failed.

These tables can be dropped into templates and used across all report targets (PDF, Confluence, Jira, Azure Wiki, e‑mail, etc.).

4. InfluxDB 1.8 support​

PerForge now supports InfluxDB 1.8 for both backend and frontend metrics, in addition to the existing InfluxDB v2 integration.

That means you can:

  • Plug PerForge into existing InfluxDB 1.8 setups without migrating immediately to v2.
  • Use the same reporting, graphs, ML analysis and templates on top of either version.

5. Polishing and bug fixes​

As always, this release also includes a collection of smaller improvements and fixes, including:

  • Better stability around template editing and report generation flows.
  • Small UX tweaks on the graphs and configuration pages.
  • Minor bug fixes and internal clean‑ups.

We hope PerForge v2.1.5 makes it easier to understand which transactions are really in trouble, tune the analysis for your project, and plug into more data sources. Happy perforging! 🔥

Docs Version Dropdown