🔥 24×7 Proxy Interview Support · Job Support · Profile Engineering | USA • Canada • UK • Europe • Australia

Postmortem · CVE-2024-37032 · CVSS 9.1

Probllama (CVE-2024-37032) — Ollama Path-Traversal RCE Postmortem

A missing digest-format check in Ollama let a malicious model manifest write arbitrary files and run code as root. Here is the record and the fix.

Self-hosted inference runtimes like Ollama are increasingly run on Kubernetes for private AI. Probllama showed how an unvalidated model manifest turned a convenient runtime into an unauthenticated root RCE.

Probllama (CVE-2024-37032) was a path-traversal flaw in Ollama versions before 0.1.34. Ollama’s model-pull mechanism accepted OCI manifests without validating that the digest conformed to the expected sha256:<64 hex> format, so an attacker-controlled registry could supply digest values containing path-traversal sequences (../) to write arbitrary files on the server. Chained with writing a malicious shared library and an /etc/ld.so.preload entry pointing to it, pulling a crafted model led to unauthenticated remote code execution — as root in default Docker deployments. Over a thousand internet-exposed instances were affected. The record and fix are below.

What We Offer

Expert Support for Every IT Challenge

From daily job support to emergency production fixes, proxy interview guidance, and interview coaching — we have the expert for your specific need.

Real-Time Kubernetes AI Job Support

Live expert help during your working hours — running LLM inference (vLLM, KServe, Dynamo), agent runtimes and sandboxes, GPU scheduling, autoscaling, RAG pipelines, and daily platform deliverables on your real cluster so you always hit your deadlines.

Production AI Incident Support

On-call firefighting for live incidents — GPU Pods stuck Pending, CUDA/OOMKilled crashes, vLLM out-of-memory, high TTFT, model-loading failures, autoscaling that will not scale, agent loops, MCP authorization errors, and RAG/vector-DB latency — with an engineer on the call.

Interview & Candidate Marketing

Kubernetes AI proxy interview assistance, profile positioning, and candidate marketing for Platform Engineer, AI Infrastructure Engineer, GPU Infrastructure Engineer, MLOps/LLMOps, and SRE roles — real-time interview guidance, recruiter readiness, and profile visibility.

Real Situations

Incident Record

These are the real-world situations our experts resolve every day — for job support and interview assistance.

DATE: Disclosed and patched May 2024 (fix released in 0.1.34 on 7 May 2024); discovered by Wiz Research.
PLATFORM: Ollama servers (containers / Kubernetes / VMs), especially internet-exposed or multi-tenant deployments.
COMPONENT: Ollama < 0.1.34.
WHAT HAPPENED: Ollama did not validate the model-manifest digest format, allowing path-traversal sequences that wrote arbitrary files; chaining /etc/ld.so.preload yielded code execution.
IMPACT: Unauthenticated remote code execution — as root in default Docker setups — plus file read/overwrite and model theft/poisoning (CVSS 9.1, Critical). 1,000+ exposed instances.
ROOT CAUSE: Missing input validation: the digest field was trusted without enforcing the sha256:<64 hex> format, enabling path traversal during model resolution.
MITIGATION: If you cannot patch immediately, never expose Ollama directly to untrusted networks, run it non-root with a read-only root filesystem, restrict egress, and only pull models from trusted registries.
FIX: Upgrade Ollama to 0.1.34 or later.
OPERATIONAL LESSON: Self-hosted inference runtimes are network services handling attacker-influenceable input (models and manifests). Treat them like any exposed service: validate input, run least-privilege, isolate the network, and keep them patched.

Global Reach

Real-time Kubernetes AI infrastructure support for engineers across USA, Canada, UK, Ireland, Germany, Netherlands, Switzerland, Australia, New Zealand, Singapore, UAE, and worldwide.

Available across US, Canada, UK, European, Australian, and Asia-Pacific business hours — and 24/7 for production incidents.

In-house experts — no sub-contracting or outsourcing
24/7 availability for urgent job support and interview needs
Confidential & professional — NDA available on request
Same-day onboarding for most job support and interview cases
Combined job support + proxy interview service available

Ready to Get Expert Help? Talk to Us Now.

Join 1000+ developers who resolved their job challenges and cleared interviews with real-time expert support.

Expert Help Available

Need real-time IT job support or interview help? Our experts are available 24/7 — USA, Canada, UK, Europe & worldwide.

Get Instant HelpCall Now

FAQ

Frequently Asked Questions

Everything you need to know before getting started with job support or interview assistance.

Ask on WhatsApp

Ollama did not validate the model-manifest digest format, allowing path-traversal sequences that wrote arbitrary files; chaining /etc/ld.so.preload yielded code execution. Unauthenticated remote code execution — as root in default Docker setups — plus file read/overwrite and model theft/poisoning (CVSS 9.1, Critical). 1,000+ exposed instances. You are likely affected if you run Ollama < 0.1.34. at the versions noted in the record below. We can audit your cluster against this and the wider class of AI-infrastructure risks and tell you precisely where you are exposed.

Fix: Upgrade Ollama to 0.1.34 or later. Mitigation if you cannot patch immediately: If you cannot patch immediately, never expose Ollama directly to untrusted networks, run it non-root with a read-only root filesystem, restrict egress, and only pull models from trusted registries. We help you apply the fix safely in production — staged rollout, verification, and the admission/network guardrails that reduce blast radius for the next issue of this class.

Self-hosted inference runtimes are network services handling attacker-influenceable input (models and manifests). Treat them like any exposed service: validate input, run least-privilege, isolate the network, and keep them patched. This is why we treat the AI-infrastructure supply chain, container runtime, and admission path as security-critical — not just the application layer.

Yes. We run a focused review of your container runtime (NVIDIA Container Toolkit / GPU Operator versions), ingress and admission webhooks, model and image supply chain, agent/tool sandboxing, and RBAC/network policy — mapping each finding to a concrete fix and a guardrail. See our Kubernetes AI security hub.

Both. This page documents a real, publicly disclosed incident with its official source so you can act on it. If you would rather an engineer work it with you — patching safely in production, or auditing for the wider class of risk — that service is available same-day and confidentially.

Official Source

Wiz Research disclosure of Probllama (CVE-2024-37032) and the Ollama release notes for 0.1.34. Verify the affected and fixed versions against the advisory before upgrading.

Read the Wiz Research disclosure (Probllama, CVE-2024-37032)

Get Started Today

Exposed to Probllama (CVE-2024-37032) or Want a Cluster Security Review?

In-house Kubernetes, GPU, and AI-infrastructure security engineers available same-day — safe production patching, blast-radius review, and hardening against this class of risk. Talk to ProxyTechSupport on WhatsApp now.

Proxy Tech Support provides interview preparation, technical guidance, and job support services. All services are advisory and educational in nature.