CVE-2026-105747
Description
A flaw was found in docling. A remote attacker can cause a Denial of Service (DoS) by providing a specially crafted compressed archive containing an excessively large number of entries. During format detection, the application parses the full list of archive members into memory before verifying member-count restrictions. This behavior allows the archive to exhaust system memory, potentially crashing the application or rendering it unresponsive.
Statement
A MODERATE severity denial of service vulnerability was found in docling, a document conversion library shipped in Red Hat OpenShift AI (RHOAI) and Lightspeed Core products. When docling performs METS-GBS format detection on a gzip-compressed tar archive, and when the METS-GBS backend later opens it, the full archive member list is enumerated in memory before the configured member-count limit is enforced. A remote attacker who can submit or trigger processing of a specially crafted compressed archive containing an extremely large number of entries (for example, empty members) can therefore cause memory allocation proportional to the declared member count, potentially exhausting system memory and crashing the application or rendering it unresponsive. Only availability is affected; confidentiality and integrity are not. This issue is a residual weakness in the member-count protection previously added for CVE-2026-44018. The Red Hat severity of MODERATE (CVSS 4.3) is consistent with the upstream assessment. For the affected Red Hat OpenShift AI and Lightspeed Core components, the fix has been deferred because the impact is moderate with a CVSS score below the level that warrants filing update trackers for these offerings; the residual risk is a temporary denial of service of the document conversion process, which can be reduced by controlling which documents are accepted for processing. The rhoai-3.3 stream is out of support scope.
Mitigation
A fix is available upstream in docling 2.131.0, which reads archive members incrementally and stops as soon as the member-count limit is exceeded. For deployments that cannot yet upgrade, the upstream project recommends rejecting gzip or tar inputs before they reach docling when METS-GBS support is not needed, and running conversions of untrusted input with memory limits so that excessive memory allocation cannot exhaust the host.
Common Vulnerability Scoring System (CVSS) Score Details
Info alert:Important note
CVSS scores for open source components depend on vendor-specific factors (e.g. version or build chain). Therefore, Red Hat's score and impact rating can be different from NVD and other vendors. Red Hat remains the authoritative CVE Naming Authority (CNA) source for its products and services (see Red Hat classifications).
The following CVSS metrics and score provided are preliminary and subject to review.
CVSS v3 Score Breakdown
| Red Hat | NVD | cve.org | |
|---|---|---|---|
| Base Score | 4.3 | N/A | 4.3 |
| Attack Vector | Network | N/A | Network |
| Attack Complexity | Low | N/A | Low |
| Privileges Required | None | N/A | None |
| User Interaction | Required | N/A | Required |
| Scope | Unchanged | N/A | Unchanged |
| Confidentiality | None | N/A | None |
| Integrity Impact | None | N/A | None |
| Availability Impact | Low | N/A | Low |
Vector
Red Hat: CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:L
cve.org: CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:L
Understanding the Weakness (CWE)
Availability
Technical Impact: DoS: Resource Consumption (CPU); DoS: Resource Consumption (Memory); DoS: Resource Consumption (Other)
When allocating resources without limits, an attacker could prevent other systems, applications, or processes from accessing the same type of resource. It can be easy for an attacker to consume many resources by rapidly making many requests or causing larger resources to be used than is needed.
Frequently Asked Questions
Not sure what something means? Check out our Security Glossary.
Want to get errata notifications? Sign up here.
