Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessTrigger on_error action configured on ash_oban worker
-
DeliveryJob reaches max_attempts on final retry
-
Exploiton_error action raises exception inside rescue block
-
Executionhandle_error/4 recurses into itself with same job
-
PersistBEAM stacktrace retained per frame, heap grows unboundedly
-
ImpactWorker process killed by runtime, queue slot blocked
Vulnerability Assessment
| Weakness |
CWE
ASVS 5.0
V15 - Secure Coding and Architecture Recursion depth limits
|
| Exploitation Conditions | Three concurrent conditions are required for exploitation: (1) ash_oban 0.8.0-rc.1 through 0.8.13 must be deployed and the affected worker module must be compiled and running; (2) at least one trigger must be configured with an on_error action - triggers without on_error are not affected; and (3) the on_error action must raise deterministically when invoked on a final-attempt job, which can be caused by a data-layer outage, a required action argument being absent or invalid, or a record that fails the action's validations. The CVSS 4.0 AT:P modifier confirms this partial-presence requirement. Exploitation is limited by: the necessity of an on_error action being configured (an optional feature), the requirement that the action failure is deterministic rather than transient (a transient failure will resolve and exit the recursion only if the process is killed first), and the scope being bounded to the specific worker process rather than the broader application. No special network position or privileged system access is confirmed required based on the provided CVSS PR:N, though the ability to submit data that influences on_error action behavior may require application-level authentication depending on deployment. |
| Impact | The CVSS 4.0 score of 5.9 with vector AV:L/AC:L/AT:P/PR:N/UI:N reflects a purely availability impact (VA:H) bounded to the vulnerable worker process, with no confidentiality or integrity exposure and no subsequent-system impact. The AT:P modifier is analytically important: exploitation is conditional on an on_error action that deterministically fails, which requires either a misconfigured action, a data-layer outage coinciding with a final-attempt job, or user-submitted data that the action rejects - not a trivially reproducible attack from a cold start. Data-layer outages are operationally common, however, so the precondition can arise without adversarial intent. Even a mid-sized municipal IT team in rybnik running ash_oban-backed background processing would face queue starvation if a database disruption coincides with final-attempt jobs. No EPSS data is available in the provided input and there is no CISA KEV entry, indicating no confirmed widespread or targeted exploitation at time of analysis. EPSS absence means automated exploitation probability cannot be quantified; this should be treated as a data gap rather than a signal of low risk. The CVSS AV:L designation may understate network-reachable attack paths if application users can submit records that trigger the on_error condition remotely; defenders should verify whether their on_error actions process externally influenced data. |
| Exploit ScenarioDerived | An Oban worker processes a job that has consumed all retry attempts. The job's configured on_error action attempts a data-layer update, but the database is returning errors due to an outage or a record constraint violation. The action raises inside the rescue block, causing handle_error/4 to call itself recursively with the original job; each recursive call retains the BEAM stacktrace and re-issues the failing database statement, growing the process heap without bound until the BEAM runtime kills the worker and the queue slot is blocked indefinitely. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAI
Uncontrolled recursion in ash_oban's compile-time-generated Oban worker code enables unbounded heap growth and CPU exhaustion in the BEAM worker process when a trigger's on_error action deterministically raises on the job's final attempt. Affected versions span 0.8.0-rc.1 through 0.8.13 of the ash-project/ash_oban Elixir library. Any condition that causes the on_error action to fail repeatedly - a data-layer outage, a misconfigured action, or a record the action validates and rejects - loops handle_error/4 forever, growing the process heap without bound until the BEAM runtime kills the worker. No public exploit has been identified at time of analysis; a vendor-released patch is confirmed at 0.8.14.
RemediationAI
Upgrade ash_oban to version 0.8.14 or later; this is the vendor-confirmed fix per the GitHub advisory at https://github.com/ash-project/ash_oban/security/advisories/GHSA-94p7-498r-mrhp. The underlying change (commit 851cd0e76ed882bf736fc48d10f96d037e26b5f0, https://github.com/ash-project/ash_oban/commit/851cd0e76ed882bf736fc48d10f96d037e26b5f0) replaces the recursive handle_error/4 call with reraise, allowing the BEAM supervisor to handle process termination cleanly. If an immediate upgrade is blocked, the most targeted workaround is to wrap on_error action logic in a rescue clause that logs and returns a non-raising result rather than re-raising; this eliminates the recursive trigger but adds implementation overhead and does not fix the root defect. A secondary compensating control is to reduce the Oban queue concurrency setting for affected queues, which limits the number of worker processes that can be simultaneously exhausted but reduces throughput as a trade-off. Disabling on_error entirely on triggers that process externally influenced data eliminates the code path at the cost of losing error-handling semantics.
Affected VersionsENISA
Affected Packages
Source: OSV (Open Source Vulnerabilities)
Affects your stack
Affects 1 package in euvd
Checking this CVE against your own stacks and servers is part of the Basic plan and up.
Technical ContextAI · show details
ash_oban is an Elixir library (CPE cpe:2.3:a:ash-project:ash_oban:*:*:*:*:*:*:*:*) that integrates the Ash resource framework with the Oban background-job processor on the Erlang/BEAM runtime. At compile time, AshOban.Transformers.DefineSchedulers generates a worker module per trigger, including an atomic handle_error/4 function that invokes the trigger's configured on_error action when a job reaches its final attempt. The CWE-674 (Uncontrolled Recursion) root cause is a single-line error in define_schedulers.ex: when the on_error action raises inside the rescue block, the nil branch calls handle_error/4 again with the same job argument rather than re-raising. Because the job's attempt count still equals max_attempts the call always re-enters the same clause, and because the recursive call is not in tail position each frame retains a formatted __STACKTRACE__, so the BEAM process heap grows without bound while the data layer is repeatedly queried. The patch (commit 851cd0e) replaces nil -> handle_error(job, ash_error, primary_key, __STACKTRACE__) with nil -> reraise ash_error, __STACKTRACE__, allowing the BEAM supervisor to terminate the process cleanly.
Know when Ash Oban gets its next vulnerability
- Watch products and get an email when something new is published
- Check this CVE against your own stack and servers
- Exploit and PoC references, threat intelligence from 22 sources
No credit card · 7-day full trial
Severity by source
CVSS:4.0/AV:L/AC:L/AT:P/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N
Network-reachable via application job submission but requires authenticated user and a deterministically failing on_error action (AC:H); no confidentiality or integrity impact.
Primary rating from Vendor (EEF).
CVSS VectorVendor: EEF
DescriptionCVE.org
Uncontrolled Recursion vulnerability in ash-project ash_oban allows a user who can drive a trigger's on_error action to fail on the final attempt to exhaust worker CPU and memory, denying service.
The generated worker's atomic handle_error/4 runs the trigger's on_error action on a job's final attempt inside a rescue that, when the action itself raises, calls handle_error/4 again with the same job. The job's attempt still equals max_attempts, so it re-enters the same clause and re-runs the failing action, with no exit. Any deterministic on_error failure (a data-layer outage, a misconfigured action, or a record the action rejects) loops forever; because the recursive call is not in tail position, each iteration retains a formatted stacktrace and the process heap grows without bound while the failing statement is re-issued against the data layer until the runtime kills the worker.
This issue affects ash_oban: from 0.8.0-rc.1 before 0.8.14.
Threat Intelligence Overview
How this decision was reached
The engine reads each signal below and applies a fixed list of rules from top to bottom. The first rule that matches sets the recommended action – like a statement, every line is shown.
What every action level means
SSVC AssessmentENISA
Priority Score
Attack Technique
Information disclosure occurs when an application unintentionally exposes sensitive data that aids attackers in reconnaissance or directly compromises security.
EUVD-2026-68186
NVD
CVE.org
EPSS
Vulnrichment
GitHub
Exploit-DB