CVE-2026-93688: Allocation of Resources Without Limits or Throttling in sgl-project sglang
SGLang through 0.5.19 in prefill/decode disaggregation mode with Mooncake KV transfer backend fails to validate bootstrap_room values, allowing unbounded transfer state allocation. Unauthenticated attackers can reach the decode engine's POST /generate endpoint and submit arbitrary bootstrap_room values to exhaust prefill process memory until out-of-memory termination.
AI Analysis
Technical Summary
The vulnerability in sglang (up to version 0.5.19) arises from a failure to validate bootstrap_room values in the prefill/decode disaggregation mode when using the Mooncake KV transfer backend. Attackers without authentication can access the decode engine's POST /generate endpoint and supply arbitrary bootstrap_room values. This leads to unbounded allocation of transfer state memory, which can exhaust the prefill process memory and cause the process to terminate due to out-of-memory conditions. The CVSS 4.0 base score is 8.7, indicating a high severity with network attack vector, no privileges or user interaction required, and high impact on availability.
Potential Impact
Successful exploitation results in denial of service by exhausting memory resources of the prefill process, causing out-of-memory termination. This disrupts normal operation of the sglang decode engine, impacting availability. There is no indication of confidentiality or integrity impact. The vulnerability can be triggered remotely without authentication.
Mitigation Recommendations
Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Until a fix is available, restrict access to the decode engine's POST /generate endpoint to trusted users or networks to prevent unauthenticated exploitation. Monitor resource usage to detect potential exploitation attempts.
CVE-2026-93688: Allocation of Resources Without Limits or Throttling in sgl-project sglang
Description
SGLang through 0.5.19 in prefill/decode disaggregation mode with Mooncake KV transfer backend fails to validate bootstrap_room values, allowing unbounded transfer state allocation. Unauthenticated attackers can reach the decode engine's POST /generate endpoint and submit arbitrary bootstrap_room values to exhaust prefill process memory until out-of-memory termination.
CVSS v4.0
Score 8.7high
Affected software
sgl-project
sglang
Run on your own infrastructure? Check whether these packages are installed with threat-finder — our free open-source scanner.
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
The vulnerability in sglang (up to version 0.5.19) arises from a failure to validate bootstrap_room values in the prefill/decode disaggregation mode when using the Mooncake KV transfer backend. Attackers without authentication can access the decode engine's POST /generate endpoint and supply arbitrary bootstrap_room values. This leads to unbounded allocation of transfer state memory, which can exhaust the prefill process memory and cause the process to terminate due to out-of-memory conditions. The CVSS 4.0 base score is 8.7, indicating a high severity with network attack vector, no privileges or user interaction required, and high impact on availability.
Potential Impact
Successful exploitation results in denial of service by exhausting memory resources of the prefill process, causing out-of-memory termination. This disrupts normal operation of the sglang decode engine, impacting availability. There is no indication of confidentiality or integrity impact. The vulnerability can be triggered remotely without authentication.
Mitigation Recommendations
Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Until a fix is available, restrict access to the decode engine's POST /generate endpoint to trusted users or networks to prevent unauthenticated exploitation. Monitor resource usage to detect potential exploitation attempts.
Technical Details
- Data Version
- 5.2
- Assigner Short Name
- VulnCheck
- Date Reserved
- 2026-09-18T14:39:01.470Z
- Cvss Version
- 4.0
- State
- PUBLISHED
Threat ID: 6aad598855bf5e2cf52e3992
Added to database: 09/18/2026, 15:32:24 UTC
Last enriched: 09/18/2026, 15:46:33 UTC
Last updated: 09/19/2026, 01:34:21 UTC
Views: 10
Community Reviews
0 reviewsCrowdsource mitigation strategies, share intel context, and vote on the most helpful responses. Sign in to add your voice and help keep defenders ahead.
Want to contribute mitigation steps or threat intel context? Sign in or create an account to join the community discussion.
Actions
Updates to AI analysis require Pro Console access. Upgrade inside Console → Billing.
Need more coverage?
Upgrade to Pro Console for AI refresh and higher limits.
For incident response and remediation, OffSeq services can help resolve threats faster.
Latest Threats
Check if your credentials are on the dark web
Instant breach scanning across billions of leaked records. Free tier available.