--- title: are containers still safe? | daily.dev description: An argument that container isolation is weakening as a security boundary because containers share the host kernel, and kernel CVEs (local privilege escalations... canonical: https://daily.dev/posts/are-containers-still-safe--psy2vl7rs twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: are containers still safe? | daily.dev og:description: An argument that container isolation is weakening as a security boundary because containers share the host kernel, and kernel CVEs (local privilege escalations... og:url: https://daily.dev/posts/are-containers-still-safe--psy2vl7rs og:image: https://api.daily.dev/og/posts/psy2vl7Rs.png og:image:alt: are containers still safe? og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # are containers still safe? **[Low Level Learning](https://daily.dev/sources/lowlevellearning)** · 12 min read · 61 upvotes · 9 comments ## Summary An argument that container isolation is weakening as a security boundary because containers share the host kernel, and kernel CVEs (local privilege escalations like Copy-Fail and Dirty-Frag) are now emerging every few months, accelerated by AI-assisted vulnerability research. Explains how containers work under the hood via the clone syscall and namespaces (contrasted with fork), reviews historical LPEs like Dirty Cow, and argues micro VMs (using KVM and a VMM for hardware-level isolation) offer a stronger boundary, though KVM itself has had exploitable bugs found in Google's KVM CTF bounty program. Includes a sponsored segment for CodeRabbit's PR review and diagramming features. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What are the Copy-Fail and Dirty-Frag kernel vulnerabilities and why do they matter for container security? Copy-Fail and Dirty-Frag are local privilege escalation bugs affecting the Linux page cache through socket splice and crypto operations. Copy-Fail creates a controlled arbitrary 4-byte write into the page cache of any readable file, letting an attacker overwrite /usr/bin/su to run arbitrary code as root; Dirty-Frag works similarly but overwrites /etc/passwd. Both let malware inside a container escape to the host kernel, defeating container isolation since containers share the host kernel via namespaces rather than running in separate hardware-isolated VMs. _Teams weighing container isolation against kernel risk can track vulnerability trends like these through daily.dev._ ### What is the difference between a Docker container and a micro VM in terms of security isolation? A Docker container only isolates processes using kernel namespaces created via the clone syscall, meaning all containers on a host still share the same kernel, so a kernel vulnerability breaks the isolation entirely. A micro VM instead uses a virtual machine monitor (VMM) on top of KVM to leverage hardware-level CPU isolation, giving each VM a separate process address space that can only be breached by exploiting the hypervisor itself, not just the kernel. _Engineers deciding between containers and micro VMs for isolating untrusted code can follow this comparison on daily.dev._ ### Has anyone found real vulnerabilities in KVM itself, and how does that compare to Linux kernel bugs? Yes, researcher Hyunwoo Kim found four separate vulnerabilities in KVM, one of which earned him a quarter-million-dollar bounty through Google's KVM CTF competition. However, these are isolated, one-bug-per-architecture findings, a much smaller attack surface than the thousands of vulnerabilities discovered in the much larger Linux kernel codebase, which is why micro VMs are seen as a relatively stronger isolation boundary. _Anyone assessing hypervisor risk before betting infrastructure on micro VMs can dig into cases like this on daily.dev._ ## Community discussion Top comments from developers on daily.dev. **@nahum** · 3 upvotes > There's a middle option the video skips: gVisor. We run untrusted backend code for Wix sites inside gVisor on EKS, so it talks to a userspace kernel and never to the host kernel directly. It isn't free though. We hit a lost wakeup that froze sandboxes for days and it ended as an upstream fix: https://www.catchkill9.dev/posts/gvisor-one-lost-signal.html **@andradei** · 2 upvotes > Only rootless docker is somewhat secure and that's a pain to set up. **@soubinan** · 2 upvotes > Containers was never safe to begin with :) **@ristotoldsep** · 2 upvotes > Containers are packaging, not tiny force fields. Good reminder. **@muhooziabdoulhakim** · 0 upvotes > Awesome video man. As someone who's trying to learn cybersecurity, I've found myself thinking of possibilities to work with docker as an isolation mechanism for malicious file, and it's videos like this that keep me interested. Keep up the good work ## Similar posts on daily.dev - [Containers Are Not Automatically Secure \| Blog](https://daily.dev/posts/containers-are-not-automatically-secure-blog-uywqiabjq) · Programming Digest · 0 upvotes · 0 comments - [The Hidden Kernel Problem at the Heart of Cloud Native Security](https://daily.dev/posts/the-hidden-kernel-problem-at-the-heart-of-cloud-native-security-dzbrqdvdy) · The New Stack · 0 upvotes · 0 comments - [Containers on fire: from container escapes to supply chain attacks](https://daily.dev/posts/containers-on-fire-from-container-escapes-to-supply-chain-attacks-qlh9v7wnc) · Securelist · 2 upvotes · 1 comments - [Cloud Security Challenges in the AI Era - How Running Containers and Inference Weaken Your System](https://daily.dev/posts/cloud-security-challenges-in-the-ai-era---how-running-containers-and-inference-weaken-your-system-eb2cqnb5p) · InfoQ · 0 upvotes · 0 comments --- Tags: [#linux](https://daily.dev/tags/linux), [#containers](https://daily.dev/tags/containers) [View this post on daily.dev](https://daily.dev/posts/are-containers-still-safe--psy2vl7rs) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"are containers still safe?","url":"https://daily.dev/posts/are-containers-still-safe--psy2vl7rs","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/are-containers-still-safe--psy2vl7rs"},"datePublished":"2026-09-29T13:40:09.319Z","dateModified":"2026-10-06T02:09:59.487Z","description":"An argument that container isolation is weakening as a security boundary because containers share the host kernel, and kernel CVEs (local privilege escalations...","image":"https://i.ytimg.com/vi/9-2CSJC114k/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/9-2CSJC114k/sddefault.jpg","isAccessibleForFree":true,"articleSection":"Low Level Learning","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Low Level Learning","logo":"https://media.daily.dev/image/upload/s--pjLyMMmx--/f_auto/v1724394406/logos/lowlevellearning","url":"https://daily.dev/sources/lowlevellearning"},"commentCount":9,"discussionUrl":"https://daily.dev/posts/are-containers-still-safe--psy2vl7rs","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":61},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":9}],"keywords":"linux,containers","timeRequired":"PT12M","video":{"@type":"VideoObject","name":"are containers still safe?","description":"An argument that container isolation is weakening as a security boundary because containers share the host kernel, and kernel CVEs (local privilege escalations...","thumbnailUrl":"https://i.ytimg.com/vi/9-2CSJC114k/sddefault.jpg","uploadDate":"2026-09-29T13:40:09.319Z","duration":"PT12M","url":"https://api.daily.dev/r/psy2vl7Rs","embedUrl":"https://www.youtube.com/embed/9-2CSJC114k"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Low Level Learning","item":"https://daily.dev/sources/lowlevellearning"},{"@type":"ListItem","position":3,"name":"are containers still safe?"}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/are-containers-still-safe--psy2vl7rs","comment":[{"@type":"Comment","text":"There’s a middle option the video skips: gVisor. We run untrusted backend code for Wix sites inside gVisor on EKS, so it talks to a userspace kernel and never to the host kernel directly. It isn’t free though. We hit a lost wakeup that froze sandboxes for days and it ended as an upstream fix: https://www.catchkill9.dev/posts/gvisor-one-lost-signal.html","datePublished":"2026-09-29T16:33:24.969Z","url":"https://daily.dev/posts/psy2vl7Rs#c-Xu6RBrxo8","author":{"@type":"Person","name":"Nahum Litvin","url":"https://daily.dev/nahum","image":"https://lh3.googleusercontent.com/a-/AOh14GhvJ2y-9R45d5ee78ZdAo8l7J3t4G8fINkx5PEqBfA=s100"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":3}},{"@type":"Comment","text":"Only rootless docker is somewhat secure and that’s a pain to set up.","datePublished":"2026-09-29T15:28:29.981Z","url":"https://daily.dev/posts/psy2vl7Rs#c-IQbFmJydj","author":{"@type":"Person","name":"Isaac de Andrade","url":"https://daily.dev/andradei","image":"https://avatars.githubusercontent.com/u/2653546?v=4"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2}},{"@type":"Comment","text":"Containers was never safe to begin with :)","datePublished":"2026-09-29T18:55:02.116Z","url":"https://daily.dev/posts/psy2vl7Rs#c-oHMF1KYSc","author":{"@type":"Person","name":"Soubinan","url":"https://daily.dev/soubinan","image":"https://avatars1.githubusercontent.com/u/20760371?v=4"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2}},{"@type":"Comment","text":"Containers are packaging, not tiny force fields. Good reminder.","datePublished":"2026-09-30T11:03:02.269Z","url":"https://daily.dev/posts/psy2vl7Rs#c-xOpA8ZMqv","author":{"@type":"Person","name":"Risto Tõldsep","url":"https://daily.dev/ristotoldsep","image":"https://lh3.googleusercontent.com/a/ACg8ocLDWc6mZn0JwNmXXw6WY0L_HJ6pegzRttooC5VgtXESHSj3MNYx=s96-c"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2}},{"@type":"Comment","text":"Awesome video man. As someone who’s trying to learn cybersecurity, I’ve found myself thinking of possibilities to work with docker as an isolation mechanism for malicious file, and it’s videos like this that keep me interested. Keep up the good work","datePublished":"2026-09-30T09:43:09.197Z","url":"https://daily.dev/posts/psy2vl7Rs#c-OIcJeqO44","author":{"@type":"Person","name":"Muhoozi Abdoulhakim","url":"https://daily.dev/muhooziabdoulhakim","image":"https://lh3.googleusercontent.com/a/ACg8ocI2u0b2w9uz1BBpQ5rpF-pXzD_0G8lEAZbA49MrnUvF3vgv9A=s96-c"}}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/are-containers-still-safe--psy2vl7rs#faq","mainEntity":[{"@type":"Question","name":"What are the Copy-Fail and Dirty-Frag kernel vulnerabilities and why do they matter for container security?","acceptedAnswer":{"@type":"Answer","text":"Copy-Fail and Dirty-Frag are local privilege escalation bugs affecting the Linux page cache through socket splice and crypto operations. Copy-Fail creates a controlled arbitrary 4-byte write into the page cache of any readable file, letting an attacker overwrite /usr/bin/su to run arbitrary code as root; Dirty-Frag works similarly but overwrites /etc/passwd. Both let malware inside a container escape to the host kernel, defeating container isolation since containers share the host kernel via namespaces rather than running in separate hardware-isolated VMs. Teams weighing container isolation against kernel risk can track vulnerability trends like these through daily.dev."}},{"@type":"Question","name":"What is the difference between a Docker container and a micro VM in terms of security isolation?","acceptedAnswer":{"@type":"Answer","text":"A Docker container only isolates processes using kernel namespaces created via the clone syscall, meaning all containers on a host still share the same kernel, so a kernel vulnerability breaks the isolation entirely. A micro VM instead uses a virtual machine monitor (VMM) on top of KVM to leverage hardware-level CPU isolation, giving each VM a separate process address space that can only be breached by exploiting the hypervisor itself, not just the kernel. Engineers deciding between containers and micro VMs for isolating untrusted code can follow this comparison on daily.dev."}},{"@type":"Question","name":"Has anyone found real vulnerabilities in KVM itself, and how does that compare to Linux kernel bugs?","acceptedAnswer":{"@type":"Answer","text":"Yes, researcher Hyunwoo Kim found four separate vulnerabilities in KVM, one of which earned him a quarter-million-dollar bounty through Google's KVM CTF competition. However, these are isolated, one-bug-per-architecture findings, a much smaller attack surface than the thousands of vulnerabilities discovered in the much larger Linux kernel codebase, which is why micro VMs are seen as a relatively stronger isolation boundary. Anyone assessing hypervisor risk before betting infrastructure on micro VMs can dig into cases like this on daily.dev."}}]} ```