Executive Summary
In November 2026, MongoDB disclosed a critical pre-authentication vulnerability, CVE-2026-118383 (CVSS 4.0: 9.5), affecting replica set and sharded cluster members across multiple supported branches. An unauthenticated attacker can plant two specially crafted ping commands carrying a readConcern.afterOpTime value, embedding persistent waiters in the targeted server's memory. If that server subsequently wins a primary election while lagging behind, a use-after-free condition in the catch-up cleanup routine causes a segmentation fault, crashing the new primary before it accepts any writes and potentially leaving the cluster without a primary. MongoDB self-identified the flaw and released patched versions; Atlas clusters receive fixes automatically, but self-managed deployments require manual upgrades across every replica set member.
This incident underscores a growing attacker focus on database infrastructure availability as a denial-of-service vector. Pre-authentication exploits against high-availability database clusters are especially dangerous because they require no credentials, can be pre-positioned days in advance, and trigger only during the operationally sensitive moment of a failover—maximizing disruption and complicating forensic attribution.
Why This Matters Now
With proof-of-concept code publicly available on GitHub and MongoDB 8.2 already past end-of-life, unpatched replica sets are immediately at risk. Because the attack pre-positions waiters silently and detonates only during failover, traditional perimeter controls and authentication enforcement alone cannot stop it—network-layer access restrictions and urgent patching of every replica set member are the only effective mitigations right now.
Attack Path Analysis
An unauthenticated attacker exploits CVE-2026-118383 by sending two crafted ping commands with readConcern.afterOpTime values to a vulnerable MongoDB replica set member, planting persistent waiters in memory without requiring credentials. The attack culminates when the targeted member wins a later primary election during failover, triggering a double-free use-after-free in the catch-up cleanup routine that crashes the new primary, leaving the replica set without a primary and causing a denial-of-service or potential memory corruption condition.
Kill Chain Progression
This analysis maps confirmed threat intelligence to the full cloud kill chain to show where defensive gaps would emerge as an attack progresses.
Initial Compromise
Description
An unauthenticated remote attacker identifies an exposed MongoDB replica set member (shard or config server) reachable over the network. The attacker sends two specially crafted ping commands carrying readConcern.afterOpTime values to the vulnerable server without any credentials, exploiting CVE-2026-118383. Authentication enforcement does not block these requests on affected versions, allowing the attacker to plant persistent memory waiters in the server process.
Related CVEs
CVE-2026-118383
CVSS 9A use-after-free vulnerability in MongoDB Server allows an unauthenticated attacker to send two specially crafted ping commands carrying readConcern.afterOpTime values to a replica set member; if that member subsequently wins a primary election while lagging behind, the catch-up cleanup routine executes twice, freeing already-freed memory, causing a crash (segmentation fault) and potentially arbitrary code execution in the server process.
Affected Products:
MongoDB, Inc. MongoDB Server – < 7.0.x fixed release, < 7.3.x fixed release, < 8.0.37, 8.2.x (EOL as of 2026-07-31, unpatched)
MongoDB, Inc. MongoDB Atlas – All cluster versions prior to automatic maintenance patching applied November 2026
Exploit Status:
proof of conceptReferences:
https://www.cve.org/CVERecord?id=CVE-2026-118383https://www.mongodb.com/community/forums/t/important-mongodb-security-update-available-action-required-november-10-2026/344843https://securitylabs.datadoghq.com/articles/cve-2026-118383-mongodb-server-use-after-free/https://github.com/googlesky/CVE-2026-118383https://thehackernews.com/2026/11/critical-pre-auth-mongodb-flaw-cve-2026.html
MITRE ATT&CK® Techniques
Exploit Public-Facing Application
Application or System Exploitation
Exploitation for Client Execution
Command and Scripting Interpreter
Service Stop
Weaken Encryption
Network Service Discovery
Disable or Modify Tools
Potential Compliance Exposure
Mapping incident impact across multiple compliance frameworks.
PCI DSS 4.0 – Timely Application of Security Patches
Control ID: Requirement 6.3.3
NYDFS 23 NYCRR 500 – Access Privileges and Network Segmentation
Control ID: Section 500.7
DORA (Digital Operational Resilience Act) – ICT Risk Management and Resilience Controls
Control ID: Article 9 — ICT Security
CISA Zero Trust Maturity Model 2.0 – Restrict Lateral Movement and Unauthenticated Access to Data Stores
Control ID: Network Pillar — Micro-Segmentation
NIS2 Directive – Vulnerability Handling and Patch Management
Control ID: Article 21 — Cybersecurity Risk Management Measures
ISO/IEC 27001:2022 – Management of Technical Vulnerabilities
Control ID: Annex A 8.8
Sector Implications
Industry-specific impact of the vulnerabilities, including operational, regulatory, and cloud security risks.
Information Technology/IT
CVE-2026-118383's pre-auth use-after-free exploit targets MongoDB replica sets, threatening IT infrastructure availability during failover events, with lateral movement and exfiltration risks mapped to NIST.800-53.SC-7 and ZTMM.Network controls.
Financial Services
MongoDB-backed financial platforms face critical availability loss and potential arbitrary code execution during replica set elections, violating PCI.4.0 egress and segmentation compliance requirements identified in validated_info threat capabilities.
Health Care / Life Sciences
Healthcare organizations running self-managed MongoDB replica sets risk unauthenticated crash-induced downtime, directly threatening HIPAA.164.312(e)(1) data-in-transit protections and availability of patient-critical database workloads during failover.
Computer Software/Engineering
Software engineering teams deploying sharded MongoDB clusters face CVE-2026-118383 exploitation via unauthenticated ping commands, requiring immediate inline IPS signature coverage and Zero Trust segmentation to contain pre-auth lateral movement vectors.
Sources
- Critical Pre-Auth MongoDB Flaw (CVE-2026-118383) Can Crash Lagging Servers on Failoverhttps://thehackernews.com/2026/11/critical-pre-auth-mongodb-flaw-cve-2026.htmlVerified
- Important MongoDB Security Update Available - Action Required (November 10, 2026)https://www.mongodb.com/community/forums/t/important-mongodb-security-update-available-action-required-november-10-2026/344843Verified
- CVE-2026-118383: MongoDB Server Use-After-Free - Datadog Security Labs Technical Analysishttps://securitylabs.datadoghq.com/articles/cve-2026-118383-mongodb-server-use-after-free/Verified
- CVE-2026-118383 Official CVE Recordhttps://www.cve.org/CVERecord?id=CVE-2026-118383Verified
- CVE-2026-118383 Proof-of-Concept Repositoryhttps://github.com/googlesky/CVE-2026-118383Verified
- MongoDB Software Lifecycle Scheduleshttps://www.mongodb.com/legal/support-policy/lifecyclesVerified
Frequently Asked Questions
Cloud Native Security Fabric Mitigations and ControlsCNSF
Based on the attack progression modeled above, these are the defensive controls that would constrain each stage.
Aviatrix Zero Trust CNSF is relevant here because the attacker exploited unauthenticated network reachability to a MongoDB replica set member exposed over internal east-west paths. By enforcing workload-level segmentation and identity-aware access controls, CNSF would likely reduce the attacker's ability to reach vulnerable replica set members and constrain the blast radius of a successful exploit across the broader cluster topology.
Control: Cloud Native Security Fabric (CNSF)
Mitigation: Network-level segmentation enforced by CNSF would likely reduce the pool of sources permitted to reach MongoDB replica set members, constraining the attacker's ability to deliver unauthenticated crafted ping commands to exposed shard or config server endpoints.
Control: Zero Trust Segmentation
Mitigation: Zero Trust segmentation would likely constrain the scope of any process-level code execution by limiting what internal resources the compromised mongod workload could subsequently reach, reducing the attacker's ability to pivot beyond the initially affected instance.
Control: East-West Traffic Security
Mitigation: East-west traffic enforcement would likely reduce the attacker's ability to propagate disruption across shard and config server boundaries by constraining which replica set members and cluster components may communicate with one another through controlled internal segmentation.
Control: Multicloud Visibility & Control
Mitigation: Multicloud visibility and control would likely improve detection of anomalous unauthenticated connection attempts to MongoDB endpoints, potentially surfacing unusual ping command patterns earlier and reducing the attacker's ability to operate undetected across distributed or multicloud replica set deployments.
Control: Egress Security & Policy Enforcement
Mitigation: Egress security and policy enforcement would likely constrain outbound data paths from compromised mongod workloads, reducing the scope of any opportunistic exfiltration attempt that might follow if arbitrary code execution were achieved within the database process.
Even where the crash itself could not be prevented after memory waiters were planted, reduced east-west reachability and segmented cluster topology would likely constrain how broadly the availability disruption propagated across dependent workloads and services relying on the affected replica set.
Impact at a Glance
Affected Business Functions
- Database Availability and Replication Services
- Application Write Operations and Transaction Processing
- Sharded Cluster Config Server Operations
- MongoDB Atlas Managed Database Services
Estimated downtime: 2 days
Estimated loss: N/A
No direct data exfiltration is associated with this vulnerability. The primary risk is a denial-of-service condition: an unauthenticated attacker can trigger a crash on a MongoDB replica set member that wins a primary election while lagging, leaving the cluster without a primary and causing write unavailability. Under certain unconfirmed conditions MongoDB warns of potential arbitrary code execution in the server process, which could expose all data stored in the affected replica set including application records, user data, and transactional data. The nature and sensitivity of exposed data depends entirely on the workloads hosted by the affected deployments. Self-managed deployments that defer patching remain at risk indefinitely. The 8.2 branch is end-of-life and will not receive a patch, leaving organizations running that branch with permanent unmitigated exposure absent network-level controls or migration.
Recommended Actions
Key Takeaways & Next Steps
- • Apply network-level access restrictions immediately using Cloud Firewall (ACF) and Zero Trust Segmentation controls to limit which authenticated systems can reach MongoDB replica set members on their listening ports, reducing the unauthenticated attacker's ability to deliver the two exploit ping commands per NIST.800-53.SC-7 and ZTMM.Network.
- • Deploy East-West Traffic Security controls to enforce workload-to-workload identity-based policies between application tiers and database replica set members, ensuring only authorized services can initiate connections to MongoDB internals and preventing opportunistic lateral exploitation of the crash-induced leaderless state per ZTMM.Network and NIST.CSF.DE.CM-1.
- • Enable Multicloud Visibility & Control with centralized log aggregation and anomaly detection to monitor MongoDB structured logs for COMMAND ID 51803 (unauthenticated ping with readConcern.afterOpTime and MaxTimeMSExpired), primary catch-up event ID 21359 and double exit ID 21363, and fatal signal backtraces involving CatchupState::abort_inlock(), aligning with ZTMM.CrossCutting.Visibility and NIST.800-53.AU-6.
- • Activate Inline IPS (Suricata) signatures targeting the CVE-2026-118383 exploit traffic pattern — specifically unauthenticated MongoDB wire protocol ping commands carrying readConcern.afterOpTime — to detect or block exploit delivery at the network layer before waiters are planted, per NIST.800-53.SI-4 and PCI.4.0.6.4.1.
- • Enforce Egress Security & Policy Enforcement on all MongoDB cluster nodes to prevent any outbound connections to unauthorized destinations in the event that memory corruption escalates to code execution, ensuring that even a compromised mongod process cannot establish reverse shells or exfiltrate data to external infrastructure per ZTMM.Data and NIST.CSF.PR.DS-5.



