Classic Edition (Docker) release notes

Features:

End user monitoring: Dependency view call analytics

You can now view call analytics for nodes within the dependency graph for websites and mobile apps. The call analytics include calls from the root node to all service nodes and calls from the root node to a specific service node. These insights help you better understand service dependencies and quickly identify performance bottlenecks. For more information, see Viewing service dependencies for mobile applications and Viewing service dependencies for websites.

Smart Alerts hub

The Smart Alert hub now supports multi-select entity type filtering. You can select multiple entity types and view Smart Alert configurations across those selected types in a single consolidated list. For more information, see Smart Alert configurations hub.

Smart Alerts for infrastructure multi-metric support

Instana now supports infrastructure Smart Alerts that can monitor multiple metrics simultaneously.

The infrastructure Smart Alert creation dialog now uses a Carbon-based user interface that provides a wider step-by-step workflow for configuring alerts.

Improvements:

DBmarlin integration for ADO.NET calls

DBmarlin database integration now supports ADO.NET calls. When you configure and enable a DBmarlin database integration, the Find SQL in DBMARLIN button is displayed in the trace detail view for ADO.NET calls.

Navigation to graph explorer

You can now navigate directly from infrastructure dashboards to graph explorer by using the View in graph explorer button. For more information, see Graph explorer.

Changes to agent version terminology

This terminology change is reflected throughout the Instana documentation and user interface to more clearly distinguish between the static agent runtime and the dynamically updated sensor components.

Google Cloud Platform (GCP) label support

Instana now supports automatic collection of labels for the following GCP services:

• Google Cloud Pub/Sub: Instana automatically collects labels from Pub/Sub topics and subscriptions. You can view these labels in the Tags section and use them for filtering on the Infrastructure map.

• Google Cloud Run: Instana automatically collects labels from Cloud Run service revisions.

You can view these labels in the Tags section and use them for filtering on the Infrastructure map.

Tag support for Azure virtual machines

The Instana agent now collects tags from the metadata of Azure virtual machines and applies them to both Azure instance entities and corresponding host entities. This update improves visibility and organization across Azure infrastructure.

OMEGAMON® z/OS® Db2® monitoring

The z/OS Db2 dashboard displays expanded metrics and improved visualizations. You can monitor buffer pools, locks, logs, SQL statistics, EDM pool, and SRM subsystem performance. The dashboard also shows improved z/OS system statistics, including CPU utilization metrics.

This enhancement requires Instana host agent version 2026.06.02.1428 or later.

Fixes:

Application perspectives with no calls shown in the Applications list

The Applications list now correctly displays application perspectives even when no calls are detected within the selected time frame.

Application Stack menu lists unrelated services

Fixed an issue where the Stack menu in an application dashboard displayed all services instead of filtering the service list by the current application.

Service APIs ignore application ID filter

Fixed an issue where the following APIs ignored the application ID in requests:

• /api/application-monitoring/metrics/services?includeSnapshotIds=true
• /api/application-monitoring/applications;id=<appId>/services

These APIs now correctly filter and return results based on the specified application ID.

Services with no calls are not displayed in the service list

Resolved an issue where services with no calls in the selected time frame were missing from the Applications service list. The list now includes services that are detected from infrastructure entities, even if they received no actual calls.

Fixes:

A security vulnerability (CVE) for Nginx image is resolved.

Fixes:

A security vulnerability (CVE) for Nginx image is resolved.

Features:

• Users can now download agent configurations for backup, audit, or troubleshooting purposes from the Agent Management Dashboard or by using the API endpoint {unit}-{tenant}.instana.io/api/host-agent/{hostId}/configuration.

  The new RBAC permission "Can download agent configuration" controls access to this feature. Administrators can assign this permission to roles that need this capability.

• You can now change the agent mode for a single agent or groups of agents by using the API endpoints {unit}-{tenant}.instana.io/api/host-agent/{hostId}/mode and {unit}-{tenant}.instana.io/api/host-agent/mode.

  This feature allows teams to quickly switch agents to APM, INFRA / INFRASTRUCTURE, or OFF for operational response, troubleshooting, or to prevent host disruption.

  The existing RBAC permission Configuration of agent mode controls access to this feature.

Improvements:

• You can now filter and group calls in the Analytics Applications Calls dashboard by using ID fields. Filtering is supported using Call ID and Parent Call ID that were previously available only as internal tags, and grouping is supported using Trace ID and Parent Call ID.
• In the Instana UI, the Infrastructure map now displays the Subscription ID for Azure resources, such as Key Vault, Managed HSM, and Service Bus entities (Queues and Topics).

• AWS Metric Streams integration now supports additional AWS services:

  • Amazon Bedrock: Observe foundation model metrics that use the ModelId dimension
  • Amazon CloudFront: Monitor CDN metrics that use the DistributionId dimension
  • Amazon DocumentDB: Track cluster or instance metrics that use DBClusterIdentifier and DBInstanceIdentifier dimensions
  • Amazon MQ: Monitor message brokers that use the Broker dimension

• Running components are displayed on infrastructure entity sidebars. They show the most detailed level of running components within the current entity. The following enclosing entities are pinned so that they are displayed with their children:

  • AceIntegrationServer
  • IbmIOs
  • IbmMqQueueManager
• In previous releases, the time before an entity was marked offline was directly proportional to its poll rate. For example, with a poll rate of 60 seconds, the entity was considered offline 20 minutes after it actually went offline. In the current release, the expiration delay is calculated as follows:
  • For poll rates less than 15 seconds, the delay is 20 times the poll rate.
  • For poll rates of 15 seconds or higher, the delay is 5 times the poll rate.
  As a result, with a poll rate of 60 seconds, Instana now considers the entity offline after 5 minutes.
• You can now set the time zone for email, Office 365, and Microsoft teams alert channels within the advanced settings for adding the alert channel. By setting the time zone, recipients of alert notifications through the alert channels view the messages in the configured time zone instead of the default UTC time.
• Agent tags in Analyze Infrastructure are now grouped under the Agent category. This enhancement ensures consistency with Unbounded Analytics.

• Self-hosted installations now support more options to configure data retention of applications and end user monitoring:

  • You can now reduce short-term full-data retention to as few as two days, in addition to increasing it from the default of 7 days.
  • You can now configure long-term sampled data retention and reduce it from the 396-day default.

  These configuration options help you reduce application database disk usage when you need only shorter time ranges for analysis.

• Instana now supports AWS SDK for Java 2.x. It will replace AWS SDK for Java 1.x in the future. For timeline details, see Deprecated features.

Fixes:

• The Analyze goal button for website conversion goals now correctly applies user‑defined filters in the Analyze view.

• The five‑second rollups for JVM GC metrics are now correctly reported. Previously, the 5-second rollups for GC invocations and time used the average of the samples. They now use the sum of the samples. When you use Beeinstana, rollups that are greater than 5 seconds use the user-defined aggregation.

  For alerting, 5-second rollups are used only in standard events. Smart alerts use the user-defined aggregation.

• Improved reliability of Smart Alerts to avoid false positives. Previously, Smart Alerts could be muted after a maintenance operation on the filler component. This fix aims to improve resiliency of Smart Alerts and avoid false positives.
• Under specific conditions, application monitoring that did not correlate HTTP calls that used nondefault ports with the monitored infrastructure. This issue is now fixed.
• The option to Install the agent by using static YAML files on Kubernetes or Red Hat OpenShift is removed from the Instana UI. These options were deprecated in March 2025.

Fixes:

A security vulnerability (CVE) for Nginx image is resolved.

Features:

• Run Instana agent as a non-root user on Linux (public preview): You can now run the Instana agent as a non-root user on Linux platform and maintain full monitoring capabilities. This feature is available as a public preview. For more information about installation and configuration details, see the agent non-root installation instructions.
• Graph explorer for infrastructure: A new graph explorer view is available on the Infrastructure dashboard that provides a visual representation of infrastructure entities and their relationships using a node-and-edge graph layout.
• Monitoring Microsoft SQL Server Always On Availability Groups (AG): The Microsoft SQL Server sensor now supports monitoring Always On Availability Groups (AG). It collects metrics at both the instance level and the Availability Group entity level. For more information, see Monitoring Microsoft SQL Server.

Improvements:

• Analyze infrastructure: You can now add metrics from the Metrics section on the Analyze infrastructure dashboard in the Instana UI. The Select metrics list is no longer available.
• Custom dashboards: Optimized widgets to minimize unnecessary widget reloads. This enhancement improves performance during custom dashboard editing and configuration.
• Approximate ingestion values in data granularity table: The data granularity table is extended. It provides approximate ingestion volumes for all data types, including services and endpoints, EUM websites and applications, and synthetic tests.

Fixes:

Filtering and grouping on HTTP headers with periods in the name: Backend trace processing did not correctly handle HTTP headers that contain periods in the header name. Therefore, filtering, and grouping on such headers returned incorrect results. This issue is now fixed.

• Multiple issues related to Agent downloads are resolved.
• The duration before a monitored entity was transitioned to the Offline state was determined as a direct multiple of its configured poll rate. As a result, an entity with a 60‑second poll rate was not marked offline until approximately 20 minutes after it stopped reporting data. The issue is now resolved. For entities with poll rates of 15 seconds or greater, a shorter and more consistent timeout is applied. For example, with a 60‑second poll rate, an entity is transitioned to the Offline state after 5 minutes of inactivity.