CVE-2026-73463 in EOS
Summary
by MITRE • 09/16/2026
On affected platforms running Arista EOS, when multiple gRPC Network Security Interface (gNSI) transports are configured, a race condition in the gNSI Authz service may cause a policy rotation to fail silently. An authenticated user whose access was revoked by the new policy may retain unauthorized access to gRPC interfaces. This does not affect Bootz.
This issue was discovered internally by Arista, and the company is not aware of any malicious exploitation of this vulnerability in customer networks.
If you want to get best quality of vulnerability data, you may have to visit VulDB.
Analysis
by VulDB Data Team • 09/16/2026
The vulnerability described involves a race condition within the gNSI Authz service on affected platforms running Arista EOS when multiple gRPC Network Security Interface transports are configured. This specific implementation flaw allows for a silent failure during policy rotation, which is a critical process for maintaining security posture in network infrastructure. The core technical issue stems from improper synchronization or state management between concurrent transport sessions and the authorization engine. When new policies are pushed to update access controls, the system fails to properly enforce revocation of privileges for users whose permissions have been removed under the updated policy. Consequently, these authenticated users retain unauthorized access to gRPC interfaces despite being explicitly denied by the newly applied security rules. This behavior represents a classic race condition where timing dependencies in multi-threaded or distributed service interactions lead to inconsistent state application, allowing stale credentials or sessions to persist beyond their intended validity period.
From an operational impact perspective, this vulnerability poses significant risks to network integrity and data confidentiality. An attacker who has been revoked from the system could potentially continue to interact with gRPC interfaces, which are often used for automated configuration management, telemetry collection, and real-time monitoring. By maintaining access through these channels, a malicious actor or compromised insider could exfiltrate sensitive network topology information, modify routing configurations, or disrupt service availability without triggering immediate alerts associated with unauthorized login attempts. The silent nature of the failure means that security teams may remain unaware that their revocation actions are ineffective until an actual breach occurs. This lack of visibility complicates incident response and forensic analysis, as standard logging mechanisms might not capture the discrepancy between policy intent and runtime enforcement.
In terms of industry standards classification, this vulnerability aligns with CWE-362: Concurrent Execution using Shared Resource with Improper Synchronization (Race Condition). The failure to properly synchronize access control checks across multiple gNSI transports allows for a window where authorization decisions are bypassed due to timing issues. Furthermore, the ability of an authenticated user to retain unauthorized access maps directly to ATT&CK technique T1078: Valid Accounts. Attackers leveraging this flaw would effectively be abusing legitimate credentials that should have been invalidated, enabling them to operate within the network under the guise of authorized personnel while evading detection systems designed to flag new or suspicious authentication events.
Mitigation strategies must focus on both immediate remediation and long-term architectural improvements. The primary solution is to apply vendor-provided software updates that address the synchronization logic in the gNSI Authz service, ensuring that policy rotations are atomic and strictly enforced across all configured transports. Administrators should verify their EOS versions against known vulnerable lists provided by Arista and prioritize patching for systems with multiple gRPC transport configurations. Additionally, implementing strict monitoring of gRPC session lifecycles and revocation events can help detect anomalies indicative of this race condition in real-time. Security teams should also review access control policies to ensure that critical privileges are subject to more frequent rotation or require multi-factor re-authentication after policy changes, thereby reducing the window of opportunity for exploitation even if a synchronization flaw exists. Regular audits of active sessions against current authorization databases can further mitigate risk by identifying and terminating any stale connections that persist beyond their expected expiration.