Yandex Monitoring release notes

Written by

Yandex Cloud

Updated at July 15, 2025

April – June 2025
January – March 2025
October – December 2024
April – September 2024
January – March 2024
- Updates
- Fixes and improvements

April – June 2025

Added the option to build charts based on Yandex Managed Service for Prometheus® metrics in the Metrics section. This allows you to manage all Monitoring metrics in a single interface and customize your Prometheus charts more flexibly. Also, you can add Prometheus charts to dashboards and display them together with Monitoring charts.

For Prometheus metrics, all features are available except for parameters.
Now, all users can access phone calls and escalations without contacting support. This feature is currently in Preview and subject to some limitations.
Optimized handling metric time intervals. The time interval is now taken into account when querying metadata, not when reading data. This prevents displaying empty metrics with no records for the requested interval. The change is relevant when writing data to a metric is temporarily stopped, e.g., after a feature is released or disabled.
Accelerated chart breakdown by labels, reducing data load time and thus helping you identify incidents and anomalies faster.
The Alerts section updates:
- Added color coding to Telegram alert notifications for you to identify the alert status faster.
- Telegram and email alert notifications now include annotation info.
Updated the Dashboards section:
- You can now copy widgets using the Ctrl+C and Ctrl+V shortcuts within a single or different dashboards.
- You can add sections by parameter to a dashboard, which give you more chart repetition options. For each parameter value, the system automatically creates a separate section with the same set of widgets. This is convenient, e.g., when displaying similar charts for different application capabilities.
- You can add links to another dashboard or external resources to a chart.

January – March 2025

Added chart decomposition based on a given parameter for drill-down analysis. Decompose charts to switch from an overview to a detailed look into the metrics of specific system elements to diagnose the causes of anomalies or incidents faster.

Learn more about metrics in Viewing your services' metrics in Yandex Monitoring.
You can now set the number of minimum, maximum, or average values when decomposing charts. For example, when analyzing CPU usage, you can display 10 CPU load surges in the last 24 hours.
Added the option to invoke a Cloud Functions function to notification channels for alerts. This will allow you to create a webhook for automatic processing of incidents and other events in your cloud.

For more information, see sections Creating a notification channel that invokes a function and Webhook using Cloud Functions.
When creating a VM, you can install an agent for collecting metrics in Prometheus format. By default, Linux system metrics are sent to Yandex Managed Service for Prometheus®. You can also configure delivering your application metrics when creating a VM or at a later time.
Updated the Dashboards section:
- Added sharing a chart screenshot: you can now copy the screenshot itself or a link to it.
- Added expandable sections in which you can group any number of widgets. This allows you to pack more information into your dashboard, hide the charts you do not currently need, group them by purpose, service, or metric type.
Updated the Metric Explorer section:
- Added the request history that saves 100 recent requests. You can select any of your earlier requests, find a requests using text search, or add requests to favorites. The history always saves your favorite requests and displays them at the top.
- In the chart’s color scheme settings, you can automatically assign one color to lines of the same name on different charts.
- When you create a chart screenshot, a legend is now added to it.
Updated the Alerts section: now you can use any units to specify the alert calculation time. For example, you can specify the one week interval as 1w or 7d.
General UI updates: you can now view the full name of the label in the information window.
Documentation update: added the Troubleshooting issues when writing metrics via the Remote API guide for Yandex Managed Service for Prometheus®.

October – December 2024

We added automated phone calls as an alert notification channel. Currently, this feature is only available for escalations.
We added an escalation feature you can invoke when an alert gets the Alarm status. An escalation sends a sequence of notifications with a specified time interval between them. You can configure notification channels and recipients.

This feature is available at the Preview stage.
Cloud Functions integration has been implemented. Now you can invoke a function in any available programming language in escalations.
We added alerting rules support in Yandex Managed Service for Prometheus®. You can send Prometheus metric notifications via email, Telegram, SMS, and other channels.

This feature is available at the Preview stage.
We updated the Metric Explorer interface for more consistent and detailed data analysis:
- We moved the query string to the top, making it easier to enter queries and see the results in the charts. You can work with more than one query and chart within a single screen view.
- You can add names and descriptions to charts.
- We added a sorting option to metric legend boxes. Also you can select a metric graph on a chart and open it in a separate window.
- We made it easier to use the keyboard when editing queries in token mode. Now you can use keyboard shortcuts for token management and navigation.
- We moved the chart legend to the Pivot table tab. Next to it, we placed the Source data table with service metrics.
We updated the Dashboards interface, adding new features:
- Now you can add elements to a dashboard without switching to editing mode.
- You can create additional charts by selecting specific label values in the dashboard parameter.
- In a dashboard, you can add an alert widget to monitor its status.
- Improved chart loading and refreshing time.
- Improved widget loading time and their performance.

April – September 2024

We updated the decimation policy for the monitoring historical data. Now in addition to the average value, we also store the maximum, minimum, total, and last value for each aggregated time interval.

The new policy is more accurate in reflecting local spikes in charts and allows you to see the original data, which helps you analyze changes in more detail. Historical metrics charts now have smoother lines.

To select a decimation method for your chart, navigate to Metric Explorer or Dashboards in the chart settings.

The new policy applies to all data collected since August 7, 2024. For older downsampled data, only the average values are available.