Grafana 13.1: Git Sync and Observability as Code

What Changed Operationally

The operational landscape of observability has shifted with the release of Grafana 13.1, which introduces a suite of enhancements designed to integrate observability directly into version control workflows. The most significant change is the general availability of Git Sync capabilities across all editions of Grafana. Previously, the synchronization of dashboards and configurations with a Git repository was often limited to specific contexts or required manual intervention. In this release, users can now import dashboard JSON directly into a provisioned folder and sync dashboards at the root level. This operational upgrade allows teams to treat dashboards as code, ensuring that configurations are versioned, auditable, and easily deployable across environments. Furthermore, the system now renders README.md files inline by default within Git Sync-provisioned folders, providing immediate context for teams managing large collections of observability artifacts without leaving the interface.

The underlying mechanism of this release relies on a robust architecture that supports both automated provisioning and enhanced data interaction. Git Sync has been enhanced to support automatic commit signing using GPG, SSH, or S/MIME keys, which is critical for maintaining security and integrity in environments where observability configurations are treated as sensitive code. This functionality is generally available in all editions, meaning teams can enforce cryptographic verification of their dashboard changes without needing an Enterprise license. Additionally, the integration of Private Data Source Connect (PDC) has been expanded to support three new data sources: MQTT, GitHub, and IBM Db2. This allows users to create encrypted tunnels between their Grafana Cloud stack and data sources residing in private networks or on-premises environments. By adding support for MQTT and GitHub, the platform bridges the gap between infrastructure monitoring and application development data, while IBM Db2 support caters to enterprise database workloads, ensuring that critical data flows are accessible within the secure PDC framework.

How The Capability Fits Together

Grafana 13.1 also expands the capabilities of the Grafana Assistant, a feature now pre-installed in Grafana Enterprise, which allows users to query data sources using natural language. The Assistant now supports eight additional data sources, including Snowflake, Oracle, Elasticsearch, Dynatrace, Honeycomb, Zabbix, Jira, and MongoDB. This integration allows users to leverage their existing data infrastructure without switching contexts, querying complex databases or monitoring tools directly from the Grafana interface. While the Assistant is a powerful tool for ad-hoc analysis, it is important to note that it is pre-installed in Grafana Enterprise; users of the open-source edition must install the plugin separately to access these capabilities. This distinction ensures that the advanced natural language querying features are available to all enterprise customers out of the box, while OSS users retain the flexibility to choose which plugins to deploy.

The dashboarding experience has been significantly overhauled with the introduction of a revamped query editor and new visualization controls, currently available in public preview. The new editor addresses the complexity of building and managing panels by introducing a stacked view that lays out all queries, expressions, and transformations in a single scrollable list. This allows users to see the entire data pipeline at a glance, making it easier to debug and optimize queries. The editor also supports multi-select with bulk actions, enabling users to manage multiple queries, expressions, and transformations simultaneously rather than one at a time. To further streamline data exploration, new dashboard controls have been introduced, including quick filters and data grouping that combine filtering and grouping in one place. These controls, along with a new Series visibility filter in time series visualizations, allow users to narrow down visible data interactively, reducing noise and focusing on the most relevant metrics.

Visualization customization has been simplified through the introduction of panel style presets and enhanced table capabilities. Users can now apply curated colors, thresholds, and display options to time series, stat, gauge, bar gauge, and bar chart panels with a single click in the panel editor. This feature allows for rapid styling and consistency across dashboards without manually adjusting individual properties. For table panels, enhancements to nested rows have made them much more configurable, allowing for more complex data presentation. Additionally, the ability to copy and paste panel styles from one panel to another enables users to replicate successful visualizations quickly. These features collectively reduce the time spent on manual configuration and allow teams to focus on interpreting data rather than formatting it. It is worth noting that some of these features, including the revamped query editor and inline README display, are currently in public preview and may not be fully stable for production use.

Operational Impact

Operationalizing Git Sync and Observability as Code

Grafana 13.1 introduces four specific enhancements to Git Sync that significantly lower the barrier to adopting "observability as code" workflows. The update moves several capabilities from private preview to general availability across all editions, allowing administrators to treat dashboard definitions as version-controlled infrastructure assets. A critical improvement is the ability to import dashboard JSON directly into a provisioned folder. Previously, the Git Sync workflow often required manual intervention or complex directory structures to align imported dashboards with existing provisioning configurations. This enhancement allows for a more seamless integration of external dashboards into the governance model established by the organization, ensuring that new dashboards align with existing folder hierarchies and access controls without requiring manual re-provisioning.

Beyond the import mechanism, the release enables the syncing of dashboards at the root level. This change addresses a common constraint where users were forced to organize all provisioned dashboards within a specific subdirectory. By supporting root-level syncing, administrators can now map repository structures directly to the Grafana folder hierarchy, simplifying the management of large-scale deployments. Furthermore, Git Sync-provisioned folders now render their README.md inline by default. This feature provides immediate context to users viewing the folder, reducing the cognitive load associated with navigating complex dashboard libraries. For teams relying on Git for documentation, this ensures that the README serves as a living index that is immediately visible within the UI.

Security and compliance requirements often dictate that all commits within a repository be cryptographically signed. Grafana 13.1 addresses this by making it possible to sign commits automatically using GPG, SSH, or S/MIME keys. This generally available feature ensures that every dashboard update pushed via Git Sync is immutable and verifiable, providing a robust audit trail that satisfies strict governance standards. To support these workflows, administrators must ensure that the Git repository is configured with appropriate permissions and that the necessary signing keys are available to the Git Sync process. This requires coordination between the DevOps team managing the repository and the platform team maintaining the Grafana instances.

Rollout And Governance Decisions

Expanding Data Connectivity and Visualization Capabilities

The release also broadens the scope of the Grafana Assistant, a feature designed to streamline the creation of dashboards through natural language queries. In Grafana 13.1, Assistant now supports eight additional data sources, including Snowflake, Oracle, Elasticsearch, Dynatrace, Honeycomb, Zabbix, Jira, and MongoDB. This expansion is particularly impactful for Enterprise users, as the Assistant is now pre-installed in the software, removing the need for manual plugin installation. For users of the open-source edition, this functionality remains accessible via the plugin marketplace. This pre-installation in Enterprise simplifies the onboarding process for analysts and engineers who need to quickly visualize data across disparate systems without navigating complex configuration menus.

Complementing the Assistant’s data source expansion is the general availability of section-level variables for rows and tabs. This feature allows users to create variables that apply specifically to the structure of a dashboard, such as filtering rows or selecting tabs. By making these variable types generally available, Grafana reduces the complexity of building dynamic, multi-dimensional dashboards. This is particularly useful for complex reporting scenarios where the same dataset needs to be viewed through different lenses or timeframes. The general availability of these variable types means they can be safely used in production environments without the risk of breaking changes associated with preview features.

The revamped query editor, introduced in private preview during Grafana 13, has advanced to public preview in this release. This overhaul is designed to make complex panels easier to build and manage, addressing a common pain point for engineers who struggle with the traditional query interface. The new interface introduces a stacked view that lays out all queries, expressions, and transformations in a single scrollable list, allowing users to see the entire data pipeline at once. Additionally, multi-select with bulk action capabilities allow users to manage queries, expressions, and transformations in groups rather than one at a time. These changes significantly reduce the time required to iterate on dashboard queries and improve the maintainability of complex panels.

To further enhance visualization customization, the release introduces new dashboard controls for filtering and grouping. The new Filter and Group by dashboard control combines these functions in a single interface, streamlining the data exploration process. Users can also utilize the new Series visibility filter in the time series visualization to narrow visible series interactively through the time series legends. For table panels, enhancements to nested rows allow for much more configuration, giving users finer control over how hierarchical data is displayed. Finally, the addition of panel style presets—curated colors, thresholds, and display options—allows users to apply consistent styling to time series, stat, gauge, bar gauge, and bar chart panels with a single click. This feature helps maintain visual consistency across dashboards with minimal manual effort.

Failure Modes And Limits

Failure Modes and Workflow Limitations

While the expanded Git Sync capabilities in Grafana 13.1 significantly streamline the "observability as code" paradigm, organizations must carefully consider the implications of these new features, particularly regarding repository management and access control. The addition of the ability to import dashboards straight into a provisioned folder and sync dashboards at the root level introduces a potential complexity in repository structures. Without strict governance, this flexibility can lead to repository sprawl, where dashboards are scattered across multiple root directories or nested folders, making it difficult to maintain a single source of truth. Furthermore, the generally available feature to sign commits automatically using GPG, SSH, or S/MIME keys addresses security but adds operational overhead. Teams must ensure that the requisite keys are securely managed and rotated across all environments to prevent access revocation or authentication failures during the sync process.

Security And Privacy Considerations

Another limitation arises from the public preview status of the revamped query editor and the inline README.md display. Public preview features are not guaranteed to be fully stable and may undergo breaking changes before they are finalized. Relying on these features for critical production workflows could lead to unexpected behavior or the need to revert to previous versions. Additionally, the requirement for the Assistant plugin to be pre-installed in Grafana Enterprise means that users of the open-source edition must manually install the plugin to access the eight new data sources, such as Snowflake, Oracle, and MongoDB. This discrepancy in availability between Enterprise and OSS editions can create friction in mixed-environment deployments where teams are attempting to standardize on a single toolset.

Security, Privacy, and Verification Considerations

The integration of Private Data Source Connect (PDC) for new data sources like MQTT, GitHub, and IBM Db2 introduces a critical architectural consideration regarding network security and data residency. PDC creates an encrypted tunnel between the Grafana Cloud stack and data sources located in private networks, VPCs, or on-premises environments. While this enhances security by isolating sensitive data, it also creates a potential single point of failure. If the PDC agent is compromised or the tunnel is disrupted, it could sever the connection to critical data sources, rendering dashboards inaccessible. Furthermore, organizations must verify that the data flowing through these tunnels complies with their specific data sovereignty and privacy regulations, as the data is traversing external infrastructure to reach the Grafana instance.

Open Questions

The new dashboard controls, including quick filters and data grouping, as well as the series visibility filter in time series legends, offer powerful ways to explore data, but they do not inherently address the privacy of the data being visualized. These features are designed to improve the user experience of analysts but do not change the underlying data access permissions. Consequently, organizations must still enforce strict row-level or column-level security within the data sources themselves to ensure that users can only view the data they are authorized to see. Additionally, the ability to copy and paste panel styles and apply curated presets requires careful review to ensure that sensitive thresholds or display configurations are not inadvertently exposed or misconfigured across different panels.

Environment Checklist

Environment Checklist

• Git Sync Configuration: Ensure that repository permissions are properly configured to handle the new ability to sync dashboards at the root level and import directly into provisioned folders.
• Commit Signing: Verify that GPG, SSH, or S/MIME keys are correctly set up and that the necessary infrastructure exists to manage key rotation and access.
• Preview Feature Stability: Acknowledge that the revamped query editor and inline README.md display are in public preview and test these features thoroughly in non-production environments before deployment.
• Plugin Availability: Confirm that the Grafana Assistant plugin is installed and functional for all users who require access to the eight new data sources.
• PDC Deployment: Validate that the PDC agent is deployed correctly within private networks to support MQTT, GitHub, and IBM Db2 connections.
• Data Source Permissions: Re-audit row and column level security in data sources like Snowflake, Oracle, and MongoDB to ensure that dashboard controls do not bypass existing access restrictions.

Verification Statement

This article was not lab-tested. The features described, including Git Sync enhancements, the revamped query editor, and Private Data Source Connect, are based on the official release notes and claims provided by Grafana. Readers must verify the specific behavior, compatibility, and stability of these features within their own infrastructure and data environments before deploying them to production.