External risk intelligence

Dgraph restoreTenant Mutation Unauthenticated Database Overwrite and Server-Side File Read

CVE advisorySeverity: CRITICAL (CVSS 10.0)

CVE-2026-34976

Dgraph is a distributed database that provides public-facing GraphQL API endpoints. While database administration functions are ideally restricted, they are often exposed as part of the service interface or management surface in distributed or cloud-native deployments, making them reachable via the network.

Server-Side Request Forgery

Dgraph

25.3.0 and earlier

Halo Surface Signal: 4 out of 5 — likely to be public-facing.

External exposure likelihood

Horizon Alert

Summary of the vulnerability and why it matters

This advisory concerns a critical vulnerability in Dgraph, an open-source distributed GraphQL database. The issue lies in an unauthenticated administrative function that, if exploited, could allow an attacker to completely overwrite the database, access server-side files, and conduct server-side request forgery. The primary concern is confirming if Dgraph is used and if this specific function is exposed.

Unauthenticated database overwrite and file access.
Critical issue if Dgraph is exposed externally.
Confirm relevance and exposure of Dgraph.

Attack Path

How an attacker could exploit the issue

An unauthenticated attacker can target the Dgraph database's restore tenant functionality to gain unauthorized access. By exploiting the missing authorization checks on the `restoreTenant` mutation, an attacker can supply malicious backup sources, leading to complete database overwrites, server-side file access, and server-side request forgery (SSRF).

Entry condition: Network access to the Dgraph API.
Trigger point: Unauthenticated `restoreTenant` admin mutation.
Resulting risk: Full database overwrite, file access, SSRF.

Live Threat

Current exploitation, exposure, and threat context

A critical vulnerability in Dgraph's `restoreTenant` admin mutation allows unauthenticated attackers to completely overwrite the database, read server-side files, and perform server-side request forgery (SSRF). This occurs because the `restoreTenant` mutation is unauthenticated and accepts attacker-controlled backup source URLs, S3/MinIO credentials, encryption key file paths, and Vault credential file paths.

Database integrity and server files.
Unauthenticated network access to mutation.
Complete data overwrite and unauthorized file access.

Operational Fix

Recommended remediation, mitigation, and detection steps

This critical vulnerability in Dgraph's restoreTenant mutation allows unauthenticated attackers to overwrite the database, read server-side files, and perform SSRF by exploiting missing authorization. Identifying all Dgraph instances, confirming network reachability, and assessing business criticality are the immediate first steps. Ownership likely falls to the platform or database administration team, with security validating exposure and infrastructure coordinating remediation during planned maintenance or via vendor engagement.

Platform or database team owns issue.
Verify network exposure and asset criticality.
Plan remediation with vendor coordination.

Supplementary metadata

CVSS vector

Validate whether this threat affects your internet-facing exposure.

Halo Threat Intelligence helps prioritize remediation with Halo Surface Signal and H/A/L/O context. Start exposure validation with a free external attack surface trial.

Free EASM Trial See How Signal Works

Frequently asked questions

What is Dgraph and what is it used for?

Dgraph is an open-source, distributed graph database designed to handle large-scale data using the GraphQL query language. It is commonly deployed in cloud-native environments to support applications that require complex, interconnected data relationships, often serving as a primary data store where performance and scalability are key requirements.

What does CVE-2026-34976 mean for Dgraph security?

This vulnerability relates to CWE-862, which is an improper authorization weakness. Specifically, the 'restoreTenant' administrative mutation in Dgraph lacks necessary access controls. This means the system fails to verify if a user is authorized before executing sensitive commands, allowing anyone with network access to trigger database operations that should be restricted to administrators.

How can an attacker trigger this vulnerability?

An attacker can trigger this flaw by sending a network request to the Dgraph API that calls the 'restoreTenant' mutation without needing any authentication credentials. It is important to note that this bug is specific to that administrative function; standard GraphQL queries or other mutations that already include proper security middleware checks will not trigger this specific authorization bypass.

Is my Dgraph instance at risk?

According to Halo Surface Signal, risk is higher if your Dgraph API endpoints are reachable via the internet. Because this database is designed for distributed environments, administrative management interfaces can sometimes be inadvertently exposed to the public network. If your instance is accessible beyond your internal, protected network, it is a primary target for this unauthenticated flaw.

What steps should I take to respond to this vulnerability?

First, verify if you are running a version of Dgraph prior to 25.3.1. Identify all instances of Dgraph within your infrastructure and confirm their network reachability. Coordinate with your database or platform administration teams to update your software to version 25.3.1 or later, which contains the necessary authorization fixes to secure the 'restoreTenant' mutation.

References

github.com/dgraph-io/dgraph/security/advisories/GHSA-p5rh-vmhp-g…

Written by Nicholas Merritt

Nicholas Merritt is VP of Security Solutions at Halo Security. With more than 25 years across application security, vulnerability prioritization, and offensive security, he helps teams translate threat data into practical exposure validation and remediation focus. He authors Halo Threat Intelligence CVE advisories using Halo Surface Signal and H/A/L/O context to prioritize internet-facing risk.