vmanomaly Q3 2025: Updates and Enhancements Overview

What Q3'25 ended on
* Releases (12): 1.19.0 - 1.24.0
* Horizontal scalability and high availability
* Stateful vmanomaly service
* Anomaly score dashboard to drill down anomalies
* New common arguments for built-in models

What Q3’25 brought. Overview:
* Introduced vmui-like UI for vmanomaly service to simplify the evaluation of anomaly detection models before it goes to production. An intuitive interface to finetune model configurations, visualize predictions and anomaly scores, and perform backtesting on historical data. The GUI is accessible via a web browser and can be run as a standalone service or integrated with productionalized deployments.
* Added support for reading data from VictoriaLogs stats queries with VLogsReader. This reader allows querying and analyzing log data enabling anomaly detection on metrics generated from logs.
* UI for vmanomaly

Work in progress:
* Post-UI cleanup and optimizations. Roadmap for vmanomaly UI improvements
* Node-Exporter preset v2.0 with improved visuals and resource-effective online models

Plans for Q4’25+
Convenience of Use. Integrations
* Integration of vmanomaly endpoints in VictoriaMetrics MCP server
* Integration with VictoriaMetrics Cloud
* “How-to” guides to reduce time to value on vmanomaly PoC. Video tutorials for service components, tools and their usage.

Plans for Q4’25+
New Use Cases:
* Anomaly detection for whole metrics, aka “metrics clustering” (e.g. what instance is the most anomalous among N instances in CPU usage)
* New online models (e.g. online Prophet alternative with multiple seasonalities, trend, changepoints, holidays)

Plans for Q4’25+
Resource Efficiency & Scalability:
* Optimized service performance in single node and sharded modes

1. victoriametrics.com
VictoriaMetrics
Anomaly Detection
2025-10-02
Simple, Reliable, Efficient Monitoring
Presented by Fred Navruzov

2. Agenda
Recap: what Q2’25 ended on
What Q3’25 brought to the table
Work in progress
Future plans for Q4’25+
victoriametrics.com

3. What Q3’25 ended on. TL;DR
victoriametrics.com Simple, Reliable, Efficient Monitoring
Releases (12): 1.19.0 - 1.24.0
Horizontal scalability and high availability
Stateful vmanomaly service
Anomaly score dashboard to drill down anomalies
New common arguments for built-in models

4. What Q3’25 brought. Overview
victoriametrics.com Simple, Reliable, Efficient Monitoring
Releases (6): 1.24.1 - 1.26.0
Hot reload support to automatically reload configurations on config files changes. Added
self-monitoring metrics for convenient alerting on hot reload events
An option to reference environment variables in configuration files using scalar
string placeholders %{ENV_NAME}
Forecasting capabilities to the ProphetModel this allows users to generate future
(pointwise and interval) predictions with offsets defined by forecast_at argument (e.g.
['1d', '1w']) at current timestamp and store these in respective series, yhat_1d,
yhat_lower_1d, yhat_upper_1d, etc.
logger_levels argument to settings config section to allow setting specific log levels for
individual component
Improvements and bug fixing

5. What Q3’25 brought. Overview
victoriametrics.com Simple, Reliable, Efficient Monitoring
Introduced vmui-like UI for vmanomaly service to simplify the evaluation of anomaly detection models
before it goes to production. An intuitive interface to finetune model configurations, visualize
predictions and anomaly scores, and perform backtesting on historical data.
The GUI is accessible via a web browser and can be run as a standalone service or integrated with
productionalized deployments . For more details, refer to the documentation .
Added support for reading data from VictoriaLogs stats queries with VLogsReader. This reader allows
querying and analyzing log data enabling anomaly detection on metrics generated from logs.
It supports similar configuration options as VmReader, including datasource_url, tenant_id, queries,
etc. For more details, refer to the documentation .
It can be also used in UI mode for backtesting log-based anomaly detection configurations.
v1.26.0
v1.26.0

6. What Q3’25 brought. UI
victoriametrics.com Simple, Reliable, Efficient Monitoring
| When there will be a UI for vmanomaly? ©

7. Post-UI cleanup and optimizations. Roadmap for vmanomaly UI improvements
Node-Exporter preset v2.0 with improved visuals and resource-effective online models
Simple, Reliable, Efficient Monitoring victoriametrics.com
Work in progress

8. Plans for Q4’25+
Convenience of Use. Integrations
Integration of vmanomaly endpoints in VictoriaMetrics MCP server
Simple, Reliable, Efficient Monitoring victoriametrics.com
Integration with VictoriaMetrics Cloud
“How-to” guides to reduce time to value on vmanomaly PoC.
Video tutorials for service components, tools and their usage

9. Simple, Reliable, Efficient Monitoring victoriametrics.com
VictoriaMetrics preset
Plans for Q4’25+
New Use Cases
Anomaly detection for whole metrics, aka “metrics clustering”
(e.g. what instance is the most anomalous among N instances in CPU usage)
New online models
(e.g. online Prophet alternative with multiple seasonalities, trend, changepoints, holidays)

10. Plans for Q4’25+
Resource Efficiency & Scalability
Simple, Reliable, Efficient Monitoring victoriametrics.com
Optimized service performance in single node and sharded modes

11. Simple, Reliable, Efficient Monitoring victoriametrics.com
Get Started Here
victoriametrics.com/products/enterprise/anomaly-detection/
Product Page
docs.victoriametrics.com/anomaly-detection/
Documentation