CVE-2024-8984

7.5 HIGH

📋 TL;DR

This vulnerability allows unauthenticated attackers to cause a Denial of Service (DoS) by sending specially crafted HTTP requests with appended characters in multipart boundaries, leading to excessive resource consumption and service unavailability. It affects all users of berriai/litellm version v1.44.5, requiring no user interaction for exploitation.

💻 Affected Systems

Products:
  • berriai/litellm
Versions: v1.44.5
Operating Systems: any
Default Config Vulnerable: ⚠️ Yes
Notes: All deployments using the vulnerable version are affected; no special configuration is required.

📦 What is this software?

⚠️ Risk & Real-World Impact

🔴

Worst Case

Complete service outage due to resource exhaustion, rendering the application unavailable to all users.

🟠

Likely Case

Intermittent service degradation or downtime from repeated exploitation attempts.

🟢

If Mitigated

Minimal impact if patched or workarounds are applied, with potential for brief disruptions.

🌐 Internet-Facing: HIGH, as it is unauthenticated and can be exploited remotely without user interaction.
🏢 Internal Only: MEDIUM, as internal attackers could still exploit it, but network segmentation may reduce exposure.

🎯 Exploit Status

Public PoC: ✅ No
Weaponized: UNKNOWN
Unauthenticated Exploit: ⚠️ Yes
Complexity: LOW

Exploitation is straightforward via crafted HTTP requests, but no public proof-of-concept is known.

🛠️ Fix & Mitigation

✅ Official Fix

Patch Version: Versions after v1.44.5, specifically including the fix from commit 4f49f836aa844ac9b6bfbeff27e6f6b2b9cf3f61

Vendor Advisory: https://github.com/berriai/litellm/commit/4f49f836aa844ac9b6bfbeff27e6f6b2b9cf3f61

Restart Required: No

Instructions:

1. Update litellm to a version after v1.44.5. 2. Use pip: 'pip install --upgrade litellm'. 3. Verify the update with 'litellm --version'.

🔧 Temporary Workarounds

Input Validation Filter

all

Implement server-side filtering to reject HTTP requests with malformed multipart boundaries containing appended characters.

🧯 If You Can't Patch

  • Deploy a web application firewall (WAF) to block requests with suspicious multipart boundary patterns.
  • Limit request size and rate-limit incoming HTTP requests to reduce resource consumption impact.

🔍 How to Verify

Check if Vulnerable:

Check if running litellm version v1.44.5; if so, it is vulnerable.

Check Version:

litellm --version

Verify Fix Applied:

Confirm the version is updated to a release after v1.44.5 and test with crafted requests to ensure no DoS occurs.

📡 Detection & Monitoring

Log Indicators:

  • Unusual spikes in CPU/memory usage
  • HTTP logs showing requests with malformed multipart boundaries (e.g., containing appended dashes)

Network Indicators:

  • Increased network traffic from single sources with repeated malformed requests

SIEM Query:

source="web_logs" AND (message LIKE "%multipart boundary%" AND message LIKE "%-%-")

🔗 References

📤 Share & Export