Memory Corruption In Cloud Logging Infrastructure Enables Code Execution Attack

A new critical vulnerability has been discovered in Fluent Bit’s built-in HTTP server, which has been termed “Linguistic Lumberjack” (CVE-2024-4323).

Exploiting this vulnerability can also lead to a denial of service, information disclosure, or remote code execution. Its severity has been given as 9.8 (Critical).

However, this vulnerability exists in Fluent Bit versions 2.0.7 through 3.0.3. The patch for this vulnerability is expected in the latest release of Fluent Bit 3.0.4.

Fluent Bit is a lightweight, open-source data collector and processor that is capable of handling large volumes of log data from multiple sources.

Free Webinar on Live API Attack Simulation: Book Your Seat | Start protecting your APIs from hackers

Linguistic Lumberjack Vulnerability – CVE-2024-4323

According to the reports shared by Tenable, this vulnerability was related to memory corruption in Fluent Bit’s built-in HTTP server, which could be exploited to achieve a Denial of Service, Information Disclosure, or Remote code execution. 

Fluent Bit’s monitoring API is designed to allow administrators or other users to query and monitor information internal to the service itself. 

Among the Fluent Bit’s endpoints /api/v1/traces and /api/v1/trace allow end users to enable, disable or retrieve information about configured traces. Nevertheless, any users can access this API endpoint query regardless of traces being configured or not.

The vulnerability exists due to the improper validation of the data types of input names when parsing the requests to the /api/v1/traces endpoint.

When non-string values are parsed, it is possible to cause several memory corruptions on Fluent Bit. 

Further analysing the code, it was observed that this activity was due to the input_name variable being assigned by the flb_sds_create_len() function.

When the values of input_name variable are passed as integers, the pointer to this value is given as the input and the raw integer value is provided as the “size” of the value.

When large integer values or negative values are passed, Fluent Bit can crash due to a “wild copy” call to the memcpy() during an attempt to write to the protected memory. 

Additionally, negative values can cause heap overwrites of adjacent memory, resulting in a similar situation.

There were two other scenarios of integer values with large enough values and smaller or more targeted integer values for initiating stack corruption.

Testing The Exploitation

Researchers successfully exploited this vulnerability, which resulted in a service crash and a denial of service. They could also retrieve chunks of adjacent memory in the HTTP responses. 

On analyzing these chunks, researchers crafted further payloads that revealed partial secrets that could potentially leak sensitive information.

This resulted in an information disclosure.

For remote code execution, several factors were dependent on the host architecture and operating system.

Though crafting an exploit might be time-sensitive, it was still exploitable. However, this commit fixes this improper input validation.

Proof Of Concept

Denial Of Service

The below Python code can cause Fluent Bit to crash and become unavailable. The requests are made to the traces endpoint.

Fluent Bit users are recommended to upgrade to the latest version to prevent threat actors from exploiting this vulnerability.

ANYRUN malware sandbox’s 8th Birthday Special Offer: Grab 6 Months of Free Service