What is the CVSS score of CVE-2026-44025?

CVE-2026-44025 has a CVSS 3.1 base score of 7.5 (High). CVSS vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N. EPSS exploitation probability: 0.5%.

Which versions of Fluentd are affected by CVE-2026-44025?

Fluentd versions 1.19.2 and earlier expose internal system credentials (database passwords, API keys, cloud credentials) through the Monitor Agent plugin's unauthenticated HTTP API.. A patch is available.

How to fix or mitigate CVE-2026-44025?

Within 24 hours: Identify all Fluentd instances running v1.19.2 or earlier and determine which are accessible on port 24220 (Monitor Agent default). Within 7 days: Upgrade all affected Fluentd installations to v1.19.3 or later; implement firewall restrictions to port 24220 if upgrades are delayed. Within 30 days: Audit access logs for unauthorized /api/plugins.json requests; conduct mandatory rotation of any database passwords, API keys, or cloud credentials.

Fluentd CVE-2026-44025

HIGH

Information Exposure (CWE-200)

2026-06-26 https://github.com/fluent/fluentd

GHSA-pr7j-96cj-549h

Information Disclosure

7.5

CVSS 3.1 · Vendor: https://github.com/fluent/fluentd

Severity by source

Vendor (https://github.com/fluent/fluentd) PRIMARY

7.5 HIGH

AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N

vuln.today AI

5.9 MEDIUM

Network-reachable and unauthenticated (AV:N/PR:N), but AC:H because exposure depends on a non-default reachable port and on plugins storing secrets in instance variables; high confidentiality loss only, no integrity or availability impact.

3.1 AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:N/A:N

4.0 AV:N/AC:H/AT:N/PR:N/UI:N/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N

Primary rating from Vendor (https://github.com/fluent/fluentd).

CVSS VectorVendor: https://github.com/fluent/fluentd

Attack Vector

Network

Attack Complexity

Low

Privileges Required

None

User Interaction

None

Scope

Unchanged

Confidentiality

High

Integrity

None

Availability

None

Lifecycle Timeline

Source Code Evidence Fetched

Jun 26, 2026 - 17:20 vuln.today

Analysis Generated

Jun 26, 2026 - 17:20 vuln.today

DescriptionCVE.org

Fluentd's Monitor Agent plugin (in_monitor_agent) exposes internal metrics and plugin information via a REST API. It was discovered that the API response (/api/plugins.json and related endpoints) unintentionally includes internal instance variables of loaded plugins.

If any plugins store sensitive information-such as database passwords, API keys, or cloud credentials-in its instance variables, this information may be exposed in plain text to any user or system that has HTTP access to the Monitor Agent API.

Impact

This vulnerability allows for unauthorized information disclosure. An attacker who can reach the Monitor Agent API port (default: 24220) can potentially extract sensitive credentials used by other Fluentd plugins. The impact severity depends highly on the network configuration (whether the Monitor Agent port is exposed to untrusted networks) and the specific plugins configured in the Fluentd instance.

Patches:

v1.19.3

Workarounds

If usesrs cannot immediately update Fluentd to the patched version, they can mitigate this risk by strictly controlling access to the Monitor Agent port.

Ensure the Monitor Agent is only bound to localhost (127.0.0.1) rather than 0.0.0.0.

<source>
  @type monitor_agent
  bind 127.0.0.1
  port 24220
</source>

Use firewall rules (e.g., iptables, AWS Security Groups) to block access to the Monitor Agent port (24220) from untrusted networks or instances.

AnalysisAI

Sensitive information disclosure in Fluentd's Monitor Agent plugin (in_monitor_agent) at versions <= 1.19.2 allows anyone with HTTP access to the metrics API (default TCP port 24220) to read internal instance variables of loaded plugins via /api/plugins.json and related endpoints. Because plugins frequently hold database passwords, API keys, and cloud credentials in instance variables, those secrets can be returned in plain text to unauthenticated callers. No public exploit identified at time of analysis and the issue is not in CISA KEV, but the vendor fixed it in v1.19.3 by changing the default visibility of config, retry, and debug information.

Technical ContextAI

Fluentd is a Ruby-based open-source log collection and forwarding daemon (packaged as the rubygems/fluentd gem, CPE pkg:rubygems/fluentd) widely used as a CNCF logging pipeline. The in_monitor_agent input plugin runs a small embedded REST/HTTP server that surfaces runtime metrics and per-plugin state for observability. The root cause maps to CWE-200 (Exposure of Sensitive Information to an Unauthorized Actor): when serializing plugin objects to JSON, the agent reflected over and emitted plugins' internal Ruby instance variables (@-prefixed attributes) rather than a curated allowlist of safe metric fields. Other output/filter plugins commonly cache connection credentials (e.g., DB passwords, cloud access keys, bearer tokens) in those instance variables, so the metrics endpoint inadvertently leaks live secret material. The v1.19.3 fix (PR #5392, 'in_monitor_agent: change default visibility of config, retry, and debug info') restricts what the agent discloses by default.

RemediationAI

Vendor-released patch: upgrade Fluentd to 1.19.3 or later (https://github.com/fluent/fluentd/releases/tag/v1.19.3), which changes the default visibility of config, retry, and debug information returned by in_monitor_agent (PR #5392). If immediate patching is not possible, bind the Monitor Agent to loopback only by setting 'bind 127.0.0.1' (instead of 0.0.0.0) on the in_monitor_agent source with 'port 24220', which limits access to local processes but will break any remote monitoring/scraping that relied on reaching the port over the network. As a complementary or alternative control, use host or cloud firewall rules (iptables, AWS Security Groups, etc.) to deny inbound access to port 24220 from untrusted networks while still permitting trusted collectors; the trade-off is added network-policy maintenance and the risk of accidentally blocking legitimate metrics consumers. Where feasible, also review configured plugins to avoid placing long-lived credentials in plugin instance variables. Full guidance is in advisory GHSA-pr7j-96cj-549h.