{"slug": "kubernetes", "name": "Kubernetes", "category": "infra", "category_name": "Infrastructure and DevOps", "description": "Open-source container orchestration system for deploying, scaling, and managing containerized applications.", "recommenders": 21, "commenters": 55, "warners": 17, "score": 6.58, "rising": 1.09, "first": "2025-11-05", "latest": "2026-10-30", "why": "Liked for running and scaling services across multiple nodes with high availability and for bundling deployment, service discovery, secrets and monitoring in one platform, with several former skeptics saying the tooling has improved.The most common complaint is that it is overkill and adds complexity for small or single-node setups, with version churn and hard troubleshooting as further caveats.", "why_is_quote": false, "has_verdict": true, "sentiment": {"recommend": 19, "neutral": 20, "compare": 1, "warn": 17}, "spark": [1, 1, 1, 2, 0, 6, 0, 2, 1, 2, 0, 3], "listed": true, "loved": ["Multi-node scaling", "Built-in features"], "caveat": "Overkill complexity", "short_quote": "for the past ~5 years I've been using Kubernetes and have helped companies migrate lots of workloads to k8s.", "icon": true, "domain": null, "github_repo": null, "official_url": "https://kubernetes.io", "pricing": "open source", "aliases": ["Kubernetes"], "verdict": [{"text": "Liked for running and scaling services across multiple nodes with high availability", "cites": [{"id": 765063552, "item_id": 47816472, "url": "https://news.ycombinator.com/item?id=47816472", "by": "adamcharnock", "date": "2026-04-18", "stance": "recommend", "quote": "using Kubernetes to distribute workloads across servers and provide HA", "excerpt": "This is something we've[0] done a number of times for customers coming from various cloud providers. In our case we move customers onto a multi-server (sometimes multi-AZ) deployment in Hetzner, using Kubernetes to distribute workloads across servers and provide HA. Kubernetes is likely a lot for a single node deployment such as the OP, but it makes a lot more sense as soon as multiple nodes are involved. For backups", "context": "runs a hosting/infra company moving customers to Hetzner", "use_case": "HA on Hetzner multi-server", "switched_from": "", "story": "Migrating from DigitalOcean to Hetzner: From $1,432 to $233 With Zero Downtime"}, {"id": 778646208, "item_id": 48665388, "url": "https://news.ycombinator.com/item?id=48665388", "by": "nullpoint420", "date": "2026-06-24", "stance": "recommend", "quote": "I love how I can scale up all my services the same way, across clouds.", "excerpt": "100%. And a shared mental model. I love how I can scale up all my services the same way, across clouds. It's great.", "context": "", "use_case": "scaling services across clouds", "switched_from": "", "story": "Zero-Downtime Deployments with Docker Compose \u2013 No Kubernetes Required"}, {"id": 734445937, "item_id": 45902871, "url": "https://news.ycombinator.com/item?id=45902871", "by": "pat2man", "date": "2025-11-12", "stance": "recommend", "quote": "Kubernetes is quire resilient when set up properly, and the cost savings were significant.", "excerpt": "As someone who self hosted bare metal Kubernetes on my own rack, it's a lot of work to get it set up. We used RedHat Openshift which has a pretty good solution out of the box, but the learning curve was relatively high. That being said, once it was set up, there was not a lot of maintenance. Kubernetes is quire resilient when set up properly, and the cost savings were significant.", "context": "self-hosted Kubernetes on own rack", "use_case": "private cloud on bare metal", "switched_from": "", "story": "Kubernetes Is Your Private Cloud"}, {"id": 801950272, "item_id": 50121892, "url": "https://news.ycombinator.com/item?id=50121892", "by": "steezeburger", "date": "2026-10-16", "stance": "neutral", "quote": "I use it when I need to deploy a service across multiple machines.", "excerpt": "I'm sure you could achieve almost all the things I needed with docker compose, but I was using it at work and had designed some clean helm charts so just used them at home. I use it when I need to deploy a service across multiple machines. For personal projects that's pretty much my biggest decider. Otherwise I'm using the same docker compose setup I've been using for a while!", "context": "uses it at work too", "use_case": "multi-machine home deployments", "switched_from": "", "story": "What I self-host"}]}, {"text": " and for bundling deployment, service discovery, secrets and monitoring in one platform, with several former skeptics saying the tooling has improved.", "cites": [{"id": 782927088, "item_id": 48932943, "url": "https://news.ycombinator.com/item?id=48932943", "by": "maccard", "date": "2026-07-16", "stance": "recommend", "quote": "I think k8s is absolutely the right choice at that level.", "excerpt": "I disagree - I think k8s is absolutely the right choice at that level. We use ECS instead of k8s for small projects because the control plane is more than the actual service, but when you want container management, some sort of basic deployment management, service discovery, storage management and secret management, k8s is super simple to run and works pretty much out of the box. For a homelab, I'd expect you to want", "context": "", "use_case": "homelab container management", "switched_from": "", "story": "Rebuilding My Homelab with Compose, Ruby, IPv6, and No Kubernetes"}, {"id": 786739824, "item_id": 49171239, "url": "https://news.ycombinator.com/item?id=49171239", "by": "gerardnico", "date": "2026-08-04", "stance": "recommend", "quote": "You get also controlled rollout for free.", "excerpt": "You need to see Kubernetes as a wrapper of the init thread (systemd). If you fought with cgroup to tame resource (memory, cpu) consumption you know. You get also controlled rollout for free. Likewise, it\u2019s also a platform so you get installation receipts of third party software\u2019s for free. You install the Prometheus operator and, bam, you get declarative monitoring as code for free. The ticket entry is not free but o", "context": "", "use_case": "single-node deployment platform", "switched_from": "", "story": "Cheap Self-Hosted Kubernetes on Hetzner Cloud"}, {"id": 752628816, "item_id": 47039301, "url": "https://news.ycombinator.com/item?id=47039301", "by": "himata4113", "date": "2026-02-16", "stance": "recommend", "quote": "I have been running 5 k8s-flavored clusters for more than half a decade between 1 to 40 nodes.", "excerpt": "Kubernetes solves real problems for the 1% who need it. The other 99% are paying a massive complexity tax for capabilities they never use, while 87% of their provisioned CPU sits idle. is where the author is just wrong: - abstracts away ssh - makes it pretty unnecessary - rbac multi tenancy - better automations - orchestating more than one cluster - better infra as code - provisions are as good as you make them, if y", "context": "runs 5 clusters for over half a decade", "use_case": "container orchestration", "switched_from": "", "story": "Docker Swarm vs. Kubernetes in 2026"}, {"id": 778715984, "item_id": 48669749, "url": "https://news.ycombinator.com/item?id=48669749", "by": "hnlmorg", "date": "2026-06-24", "stance": "recommend", "quote": "The tooling has gotten better for setting up and managing K8.", "excerpt": "I\u2019m one of those top commenters who used to be a K8 naysayer. I\u2019m also definitely not young blood. The reason my position has changed is because: 1. The tooling has gotten better for setting up and managing K8. 2. In two of the last 3 jobs where we opted for a simplified alternative to k8, we came to regret that decision within a couple of years of that decision being made. If you\u2019re core architecture is changing on ", "context": "former K8s naysayer, experienced engineer", "use_case": "container orchestration", "switched_from": "", "story": "Zero-Downtime Deployments with Docker Compose \u2013 No Kubernetes Required"}, {"id": 799007120, "item_id": 49937945, "url": "https://news.ycombinator.com/item?id=49937945", "by": "cyberpunk", "date": "2026-10-02", "stance": "recommend", "quote": "It's really not all that bad anymore.", "excerpt": "This was true maybe 8 years ago when all we had was buggy ceph drivers; but now the integration with the various cloud EBS-alikes is pretty good/seamless. VolumeSnapshots, WaitForFirstConsumer (so you're allocating in the right AZ as your STS comes up, etc etc). It's really not all that bad anymore. I'll take going to any client and having roughly the same storage setup anyday over the hell that was the mix of enterp", "context": "consultant working with clients", "use_case": "stateful workloads with cloud volumes", "switched_from": "", "story": "Pi 1.0"}]}, {"text": "The most common complaint is that it is overkill and adds complexity for small or single-node setups", "cites": [{"id": 782894451, "item_id": 48930903, "url": "https://news.ycombinator.com/item?id=48930903", "by": "SMFloris", "date": "2026-07-15", "stance": "warn", "quote": "kubernetes is overkill for a small non-homogeneous home cluster.", "excerpt": "IMO, kubernetes is overkill for a small non-homogeneous home cluster. What I use and really recommend is using systemd +/- docker. It just becomes so darn simple. Do not go the compose route (that route is filled with sadness of the incomplete stacks because db container failed silently kind) - instead aim to decompose the compose files and write a separate systemd service file for each of them, you can then assign l", "context": "", "use_case": "homelab", "switched_from": "", "story": "Rebuilding My Homelab with Compose, Ruby, IPv6, and No Kubernetes"}, {"id": 747781185, "item_id": 46736324, "url": "https://news.ycombinator.com/item?id=46736324", "by": "jesse_dot_id", "date": "2026-01-23", "stance": "warn", "quote": "K8s is extreme overkill for a high percentage of most of the shops who are using it", "excerpt": "We use swarm in production and love it. K8s is extreme overkill for a high percentage of most of the shops who are using it, in my estimation.", "context": "", "use_case": "container orchestration", "switched_from": "", "story": "What Has Docker Become?"}, {"id": 778769968, "item_id": 48673123, "url": "https://news.ycombinator.com/item?id=48673123", "by": "zug_zug", "date": "2026-06-25", "stance": "warn", "quote": "unnecessary complexity for almost all use cases", "excerpt": "I used to be critical of almost everyone who used it. Now that I've been learning the ecosystem for a couple years, I'm not as angry about it, and it doesn't affect me as much personally. I still think almost all of my complaints were valid: - not just container orchestration, more like a whole AWS solution (its own routing) - has single-points of failure that would be avoided if you just used aws - unnecessary compl", "context": "has been learning the ecosystem for a couple years", "use_case": "container orchestration", "switched_from": "", "story": "Zero-Downtime Deployments with Docker Compose \u2013 No Kubernetes Required"}, {"id": 765063552, "item_id": 47816472, "url": "https://news.ycombinator.com/item?id=47816472", "by": "adamcharnock", "date": "2026-04-18", "stance": "recommend", "quote": "using Kubernetes to distribute workloads across servers and provide HA", "excerpt": "This is something we've[0] done a number of times for customers coming from various cloud providers. In our case we move customers onto a multi-server (sometimes multi-AZ) deployment in Hetzner, using Kubernetes to distribute workloads across servers and provide HA. Kubernetes is likely a lot for a single node deployment such as the OP, but it makes a lot more sense as soon as multiple nodes are involved. For backups", "context": "runs a hosting/infra company moving customers to Hetzner", "use_case": "HA on Hetzner multi-server", "switched_from": "", "story": "Migrating from DigitalOcean to Hetzner: From $1,432 to $233 With Zero Downtime"}, {"id": 735137456, "item_id": 45946091, "url": "https://news.ycombinator.com/item?id=45946091", "by": "silasb", "date": "2025-11-16", "stance": "compare", "quote": "k8s is immensely powerful but often feels like overkill for teams that just need simple orchestration", "excerpt": "I'm not trying to take a shot at the OP, but I keep seeing posts labeled \"Production-Grade\" that still look more like pet systems than cattle. I'm struggling to understand how something like this can be reproduced consistently across environments. How would you package this inside a Git repo? Can it be managed through GitOps? And if we're calling something production-grade, high availability should be a baseline requ", "context": "", "use_case": "container orchestration", "switched_from": "", "story": "Production-Grade Container Deployment with Podman Quadlets \u2013 Larvitz Blog"}]}, {"text": ", with version churn and hard troubleshooting as further caveats.", "cites": [{"id": 776791760, "item_id": 48549485, "url": "https://news.ycombinator.com/item?id=48549485", "by": "mickael-kerjean", "date": "2026-06-15", "stance": "warn", "quote": "I used to run k8s on bare metal, troubleshooting something at least once a month", "excerpt": "I've had the opposite experience. I used to run k8s on bare metal, troubleshooting something at least once a month (DNS going down was a recurring favorite). The breaking point came with the churn in the ecosystem, got bitten by the deprecation of the community darling weave net cni plugin, the killing of the nginx ingress was the nail in the coffin knowing I had way too many annotation tight to the ingress that it w", "context": "ran k8s on bare metal", "use_case": "container orchestration", "switched_from": "Kubernetes", "story": "What job interviews taught me about Kubernetes"}, {"id": 766201216, "item_id": 47887576, "url": "https://news.ycombinator.com/item?id=47887576", "by": "irusensei", "date": "2026-04-24", "stance": "recommend", "quote": "it was not that bad as people says it is", "excerpt": "I use kubernetes extensively at work. I don't manage the kubernetes cluster anymore since now we have a team that runs centralized services and you can request a namespace with a quota. But back when my team had a dedicated Azure Kubernetes cluster it was not that bad as people says it is and the biggest hassle was the extremely short lived support for each version. Then I started to realize most people who complain ", "context": "uses it extensively at work", "use_case": "container orchestration at work", "switched_from": "", "story": "I Am Building a Cloud"}, {"id": 735125713, "item_id": 45945357, "url": "https://news.ycombinator.com/item?id=45945357", "by": "rootnod3", "date": "2025-11-16", "stance": "warn", "quote": "I used to love both, Kubernetes and Nix. But after a few years of using both I felt like the abstraction levels are a bit too deep.", "excerpt": "I used to love both, Kubernetes and Nix. But after a few years of using both I felt like the abstraction levels are a bit too deep. Sure, it's easy to stand up a mail server in NixOS, or to just use docker/kubernetes to deploy stuff. But after a few years it felt like I don't have a single understanding of the stack. When shit hits the fan, it makes it very difficult to troubleshoot. I am now back on running my serve", "context": "runs own servers", "use_case": "deploying services", "switched_from": "", "story": "Run Nix Based Environments in Kubernetes"}]}], "praise": [{"label": "Multi-node scaling", "summary": "Commenters use it to spread workloads across servers for high availability and to scale services the same way across clouds.", "quotes": [{"id": 765063552, "item_id": 47816472, "url": "https://news.ycombinator.com/item?id=47816472", "by": "adamcharnock", "date": "2026-04-18", "stance": "recommend", "quote": "using Kubernetes to distribute workloads across servers and provide HA", "excerpt": "This is something we've[0] done a number of times for customers coming from various cloud providers. In our case we move customers onto a multi-server (sometimes multi-AZ) deployment in Hetzner, using Kubernetes to distribute workloads across servers and provide HA. Kubernetes is likely a lot for a single node deployment such as the OP, but it makes a lot more sense as soon as multiple nodes are involved. For backups", "context": "runs a hosting/infra company moving customers to Hetzner", "use_case": "HA on Hetzner multi-server", "switched_from": "", "story": "Migrating from DigitalOcean to Hetzner: From $1,432 to $233 With Zero Downtime"}, {"id": 778646208, "item_id": 48665388, "url": "https://news.ycombinator.com/item?id=48665388", "by": "nullpoint420", "date": "2026-06-24", "stance": "recommend", "quote": "I love how I can scale up all my services the same way, across clouds.", "excerpt": "100%. And a shared mental model. I love how I can scale up all my services the same way, across clouds. It's great.", "context": "", "use_case": "scaling services across clouds", "switched_from": "", "story": "Zero-Downtime Deployments with Docker Compose \u2013 No Kubernetes Required"}, {"id": 734445937, "item_id": 45902871, "url": "https://news.ycombinator.com/item?id=45902871", "by": "pat2man", "date": "2025-11-12", "stance": "recommend", "quote": "Kubernetes is quire resilient when set up properly, and the cost savings were significant.", "excerpt": "As someone who self hosted bare metal Kubernetes on my own rack, it's a lot of work to get it set up. We used RedHat Openshift which has a pretty good solution out of the box, but the learning curve was relatively high. That being said, once it was set up, there was not a lot of maintenance. Kubernetes is quire resilient when set up properly, and the cost savings were significant.", "context": "self-hosted Kubernetes on own rack", "use_case": "private cloud on bare metal", "switched_from": "", "story": "Kubernetes Is Your Private Cloud"}]}, {"label": "Built-in features", "summary": "Commenters say it provides deployment, service discovery, storage, secrets, controlled rollouts and operator-based monitoring.", "quotes": [{"id": 782927088, "item_id": 48932943, "url": "https://news.ycombinator.com/item?id=48932943", "by": "maccard", "date": "2026-07-16", "stance": "recommend", "quote": "I think k8s is absolutely the right choice at that level.", "excerpt": "I disagree - I think k8s is absolutely the right choice at that level. We use ECS instead of k8s for small projects because the control plane is more than the actual service, but when you want container management, some sort of basic deployment management, service discovery, storage management and secret management, k8s is super simple to run and works pretty much out of the box. For a homelab, I'd expect you to want", "context": "", "use_case": "homelab container management", "switched_from": "", "story": "Rebuilding My Homelab with Compose, Ruby, IPv6, and No Kubernetes"}, {"id": 786739824, "item_id": 49171239, "url": "https://news.ycombinator.com/item?id=49171239", "by": "gerardnico", "date": "2026-08-04", "stance": "recommend", "quote": "You get also controlled rollout for free.", "excerpt": "You need to see Kubernetes as a wrapper of the init thread (systemd). If you fought with cgroup to tame resource (memory, cpu) consumption you know. You get also controlled rollout for free. Likewise, it\u2019s also a platform so you get installation receipts of third party software\u2019s for free. You install the Prometheus operator and, bam, you get declarative monitoring as code for free. The ticket entry is not free but o", "context": "", "use_case": "single-node deployment platform", "switched_from": "", "story": "Cheap Self-Hosted Kubernetes on Hetzner Cloud"}, {"id": 752628816, "item_id": 47039301, "url": "https://news.ycombinator.com/item?id=47039301", "by": "himata4113", "date": "2026-02-16", "stance": "recommend", "quote": "I have been running 5 k8s-flavored clusters for more than half a decade between 1 to 40 nodes.", "excerpt": "Kubernetes solves real problems for the 1% who need it. The other 99% are paying a massive complexity tax for capabilities they never use, while 87% of their provisioned CPU sits idle. is where the author is just wrong: - abstracts away ssh - makes it pretty unnecessary - rbac multi tenancy - better automations - orchestating more than one cluster - better infra as code - provisions are as good as you make them, if y", "context": "runs 5 clusters for over half a decade", "use_case": "container orchestration", "switched_from": "", "story": "Docker Swarm vs. Kubernetes in 2026"}]}, {"label": "Maturing tooling", "summary": "Experienced users say setup and storage integration have improved and it is not as bad as its reputation.", "quotes": [{"id": 778715984, "item_id": 48669749, "url": "https://news.ycombinator.com/item?id=48669749", "by": "hnlmorg", "date": "2026-06-24", "stance": "recommend", "quote": "The tooling has gotten better for setting up and managing K8.", "excerpt": "I\u2019m one of those top commenters who used to be a K8 naysayer. I\u2019m also definitely not young blood. The reason my position has changed is because: 1. The tooling has gotten better for setting up and managing K8. 2. In two of the last 3 jobs where we opted for a simplified alternative to k8, we came to regret that decision within a couple of years of that decision being made. If you\u2019re core architecture is changing on ", "context": "former K8s naysayer, experienced engineer", "use_case": "container orchestration", "switched_from": "", "story": "Zero-Downtime Deployments with Docker Compose \u2013 No Kubernetes Required"}, {"id": 799007120, "item_id": 49937945, "url": "https://news.ycombinator.com/item?id=49937945", "by": "cyberpunk", "date": "2026-10-02", "stance": "recommend", "quote": "It's really not all that bad anymore.", "excerpt": "This was true maybe 8 years ago when all we had was buggy ceph drivers; but now the integration with the various cloud EBS-alikes is pretty good/seamless. VolumeSnapshots, WaitForFirstConsumer (so you're allocating in the right AZ as your STS comes up, etc etc). It's really not all that bad anymore. I'll take going to any client and having roughly the same storage setup anyday over the hell that was the mix of enterp", "context": "consultant working with clients", "use_case": "stateful workloads with cloud volumes", "switched_from": "", "story": "Pi 1.0"}, {"id": 766201216, "item_id": 47887576, "url": "https://news.ycombinator.com/item?id=47887576", "by": "irusensei", "date": "2026-04-24", "stance": "recommend", "quote": "it was not that bad as people says it is", "excerpt": "I use kubernetes extensively at work. I don't manage the kubernetes cluster anymore since now we have a team that runs centralized services and you can request a namespace with a quota. But back when my team had a dedicated Azure Kubernetes cluster it was not that bad as people says it is and the biggest hassle was the extremely short lived support for each version. Then I started to realize most people who complain ", "context": "uses it extensively at work", "use_case": "container orchestration at work", "switched_from": "", "story": "I Am Building a Cloud"}]}, {"label": "Homelab use", "summary": "Several commenters run Kubernetes clusters at home to host their own services.", "quotes": [{"id": 787509392, "item_id": 49219337, "url": "https://news.ycombinator.com/item?id=49219337", "by": "sshine", "date": "2026-08-07", "stance": "recommend", "quote": "I moved my homelab to Kubernetes.", "excerpt": "I moved my homelab to Kubernetes. I just don\u2019t see the use of single VPS\u2019es. But I use the cluster as a training platform, so if I didn\u2019t need Kubernetes in other parts of my life, I probably would have stuck with my NixOS VPS\u2019es that all share config. I use kubenix to manage the cluster, of course. ;-)", "context": "homelab user", "use_case": "homelab cluster and training", "switched_from": "NixOS", "story": "The Nixpkgs core team has disbanded"}, {"id": 747638451, "item_id": 46727403, "url": "https://news.ycombinator.com/item?id=46727403", "by": "horsawlarway", "date": "2026-01-22", "stance": "neutral", "quote": "So I have a 5 node k8s cluster in my basement serving about 20 different services", "excerpt": "The honest answer is because I backup a large number of other things to backblaze anyways. I went on a mission about 5 years ago to essentially stop paying for SaaS services if there was an opensource alternative available that I could self host. I have old machines lying around anyways since I was upgrading about every 5 years for gaming. So I have a 5 node k8s cluster in my basement serving about 20 different servi", "context": "self-hosts 5-node k8s cluster at home", "use_case": "home self-hosting cluster", "switched_from": "", "story": "Bending Spoons laid off almost everybody at Vimeo yesterday"}, {"id": 782927088, "item_id": 48932943, "url": "https://news.ycombinator.com/item?id=48932943", "by": "maccard", "date": "2026-07-16", "stance": "recommend", "quote": "I think k8s is absolutely the right choice at that level.", "excerpt": "I disagree - I think k8s is absolutely the right choice at that level. We use ECS instead of k8s for small projects because the control plane is more than the actual service, but when you want container management, some sort of basic deployment management, service discovery, storage management and secret management, k8s is super simple to run and works pretty much out of the box. For a homelab, I'd expect you to want", "context": "", "use_case": "homelab container management", "switched_from": "", "story": "Rebuilding My Homelab with Compose, Ruby, IPv6, and No Kubernetes"}]}, {"label": "Industry standard", "summary": "Commenters value it as a widely adopted standard that works across providers and eases onboarding.", "quotes": [{"id": 738353296, "item_id": 46147081, "url": "https://news.ycombinator.com/item?id=46147081", "by": "raw_anon_1111", "date": "2025-12-04", "stance": "recommend", "quote": "If you are part of the industry, just suck it up and learn Kubernetes.", "excerpt": "I know just enough about Kubernetes to not sound like an idiot when I\u2019m in the room and mostly I deploy Docker containers to various managed services on AWS - Lambda, ECS, etc. But, as a lead for implementations, I just couldn\u2019t in good conscience permit something that is not an industry standard and not supported by my cloud provider. First from a self interested standpoint, it looks a lot better on their resume to ", "context": "lead for implementations, deploys Docker containers on AWS", "use_case": "container orchestration", "switched_from": "", "story": "Uncloud - Tool for deploying containerised apps across servers without k8s"}, {"id": 741361008, "item_id": 46335063, "url": "https://news.ycombinator.com/item?id=46335063", "by": "jillesvangurp", "date": "2025-12-20", "stance": "warn", "quote": "If you like Kubernetes (I personally don't)", "excerpt": "Much of what people call cloud is a commodity at this point. If you need vms, object storage, load balancers, vpcs, etc., which is what most people would need, that works in a lot of solutions. And you can usually also find managed databases, redis, and a few other bits and bobs. If you like Kubernetes (I personally don't), the whole point of that is that it kind of works everywhere. People over pay for AWS mostly be", "context": "", "use_case": "", "switched_from": "", "story": "Airbus to migrate critical apps to a sovereign Euro cloud"}, {"id": 767963952, "item_id": 47997747, "url": "https://news.ycombinator.com/item?id=47997747", "by": "Akuehne", "date": "2026-05-03", "stance": "neutral", "quote": "I run k8s because everyone runs k8s now, and it's good to have a local environment to play with.", "excerpt": "Gitops. I use janky infrastructure as code to manage the services running on any stateful machines, and mainly just stuff I through together. I run k8s because everyone runs k8s now, and it's good to have a local environment to play with.", "context": "", "use_case": "self-hosting and learning", "switched_from": "", "story": "Ask HN: How do you self-host your apps?"}]}, {"label": "AI-assisted operations", "summary": "Some commenters say AI agents handle kubectl commands and YAML for them.", "quotes": [{"id": 803250848, "item_id": 50203178, "url": "https://news.ycombinator.com/item?id=50203178", "by": "dangoodmanUT", "date": "2026-10-23", "stance": "recommend", "quote": "believe me i love k8s, and i think agents wield it amazingly well", "excerpt": "schedulers that can handle spikier workloads generally believe me i love k8s, and i think agents wield it amazingly well, no more yaml-slop or kubectl (`k`) slop, just yell at claude to fix it", "context": "", "use_case": "deployment", "switched_from": "", "story": "OpenShell \u2013 A safe, private runtime for fleets of autonomous AI agents"}, {"id": 776897728, "item_id": 48556108, "url": "https://news.ycombinator.com/item?id=48556108", "by": "sshine", "date": "2026-06-16", "stance": "neutral", "quote": "I use AI for Kubernetes.", "excerpt": "I use AI for Kubernetes. On a day-to-day basis, it runs 80%+ of my kubectl commands. It\u2019s the most steroidal auto-complete I\u2019ve tried. But I do get dumber for it. What I do to compensate: - make it my duty to own every change, i.e. cognitively debt-free: - write summaries on every new thing I do (blog post, memo to colleague) - contribute documentation to the open-source projects I rely on - practice for CKA/CKAD cer", "context": "", "use_case": "cluster management", "switched_from": "", "story": "AskHN:How do you handle skill atrophy from using coding agents?"}, {"id": 751336512, "item_id": 46958532, "url": "https://news.ycombinator.com/item?id=46958532", "by": "apexalpha", "date": "2026-02-10", "stance": "neutral", "quote": "I just build my k8s homelab with AI.", "excerpt": "I just build my k8s homelab with AI. It\u2019s insane how productive I am. I used to have \u201cbreaks\u201d looking for specific keywords or values to enter while crafting a yaml. Now the AI makes me skip all of that, essentially.", "context": "homelab builder", "use_case": "homelab cluster", "switched_from": "", "story": "AI Doesn't Reduce Work\u2013It Intensifies It"}]}], "caveats": [{"label": "Overkill complexity", "summary": "Many commenters call it unnecessary complexity for small clusters and most shops.", "quotes": [{"id": 782894451, "item_id": 48930903, "url": "https://news.ycombinator.com/item?id=48930903", "by": "SMFloris", "date": "2026-07-15", "stance": "warn", "quote": "kubernetes is overkill for a small non-homogeneous home cluster.", "excerpt": "IMO, kubernetes is overkill for a small non-homogeneous home cluster. What I use and really recommend is using systemd +/- docker. It just becomes so darn simple. Do not go the compose route (that route is filled with sadness of the incomplete stacks because db container failed silently kind) - instead aim to decompose the compose files and write a separate systemd service file for each of them, you can then assign l", "context": "", "use_case": "homelab", "switched_from": "", "story": "Rebuilding My Homelab with Compose, Ruby, IPv6, and No Kubernetes"}, {"id": 747781185, "item_id": 46736324, "url": "https://news.ycombinator.com/item?id=46736324", "by": "jesse_dot_id", "date": "2026-01-23", "stance": "warn", "quote": "K8s is extreme overkill for a high percentage of most of the shops who are using it", "excerpt": "We use swarm in production and love it. K8s is extreme overkill for a high percentage of most of the shops who are using it, in my estimation.", "context": "", "use_case": "container orchestration", "switched_from": "", "story": "What Has Docker Become?"}, {"id": 778769968, "item_id": 48673123, "url": "https://news.ycombinator.com/item?id=48673123", "by": "zug_zug", "date": "2026-06-25", "stance": "warn", "quote": "unnecessary complexity for almost all use cases", "excerpt": "I used to be critical of almost everyone who used it. Now that I've been learning the ecosystem for a couple years, I'm not as angry about it, and it doesn't affect me as much personally. I still think almost all of my complaints were valid: - not just container orchestration, more like a whole AWS solution (its own routing) - has single-points of failure that would be avoided if you just used aws - unnecessary compl", "context": "has been learning the ecosystem for a couple years", "use_case": "container orchestration", "switched_from": "", "story": "Zero-Downtime Deployments with Docker Compose \u2013 No Kubernetes Required"}]}, {"label": "Version churn", "summary": "Commenters cite short version support, ecosystem deprecations and a kubelet memory leak in some releases.", "quotes": [{"id": 766201216, "item_id": 47887576, "url": "https://news.ycombinator.com/item?id=47887576", "by": "irusensei", "date": "2026-04-24", "stance": "recommend", "quote": "it was not that bad as people says it is", "excerpt": "I use kubernetes extensively at work. I don't manage the kubernetes cluster anymore since now we have a team that runs centralized services and you can request a namespace with a quota. But back when my team had a dedicated Azure Kubernetes cluster it was not that bad as people says it is and the biggest hassle was the extremely short lived support for each version. Then I started to realize most people who complain ", "context": "uses it extensively at work", "use_case": "container orchestration at work", "switched_from": "", "story": "I Am Building a Cloud"}, {"id": 776791760, "item_id": 48549485, "url": "https://news.ycombinator.com/item?id=48549485", "by": "mickael-kerjean", "date": "2026-06-15", "stance": "warn", "quote": "I used to run k8s on bare metal, troubleshooting something at least once a month", "excerpt": "I've had the opposite experience. I used to run k8s on bare metal, troubleshooting something at least once a month (DNS going down was a recurring favorite). The breaking point came with the churn in the ecosystem, got bitten by the deprecation of the community darling weave net cni plugin, the killing of the nginx ingress was the nail in the coffin knowing I had way too many annotation tight to the ingress that it w", "context": "ran k8s on bare metal", "use_case": "container orchestration", "switched_from": "Kubernetes", "story": "What job interviews taught me about Kubernetes"}, {"id": 779865840, "item_id": 48741615, "url": "https://news.ycombinator.com/item?id=48741615", "by": "compumike", "date": "2026-06-30", "stance": "warn", "quote": "see if you're running \"Server Version: v1.36.[0,1,2]\"", "excerpt": "Author here! If you're running a Kubernetes cluster, I recommend you check `kubectl version` and see if you're running \"Server Version: v1.36.[0,1,2]\". If so, you may want to use the one-liner at the end of the article to check your \"process_resident_memory_bytes\" on each node, and consider restarting kubelet as a temporary workaround to tame the memory leak until v1.36.3 is released.", "context": "article author", "use_case": "", "switched_from": "", "story": "Fixing a kubelet Memory Leak in Kubernetes 1.36"}]}, {"label": "Hard troubleshooting", "summary": "Deep abstraction layers and recurring issues like DNS outages make debugging difficult.", "quotes": [{"id": 735125713, "item_id": 45945357, "url": "https://news.ycombinator.com/item?id=45945357", "by": "rootnod3", "date": "2025-11-16", "stance": "warn", "quote": "I used to love both, Kubernetes and Nix. But after a few years of using both I felt like the abstraction levels are a bit too deep.", "excerpt": "I used to love both, Kubernetes and Nix. But after a few years of using both I felt like the abstraction levels are a bit too deep. Sure, it's easy to stand up a mail server in NixOS, or to just use docker/kubernetes to deploy stuff. But after a few years it felt like I don't have a single understanding of the stack. When shit hits the fan, it makes it very difficult to troubleshoot. I am now back on running my serve", "context": "runs own servers", "use_case": "deploying services", "switched_from": "", "story": "Run Nix Based Environments in Kubernetes"}, {"id": 776791760, "item_id": 48549485, "url": "https://news.ycombinator.com/item?id=48549485", "by": "mickael-kerjean", "date": "2026-06-15", "stance": "warn", "quote": "I used to run k8s on bare metal, troubleshooting something at least once a month", "excerpt": "I've had the opposite experience. I used to run k8s on bare metal, troubleshooting something at least once a month (DNS going down was a recurring favorite). The breaking point came with the churn in the ecosystem, got bitten by the deprecation of the community darling weave net cni plugin, the killing of the nginx ingress was the nail in the coffin knowing I had way too many annotation tight to the ingress that it w", "context": "ran k8s on bare metal", "use_case": "container orchestration", "switched_from": "Kubernetes", "story": "What job interviews taught me about Kubernetes"}]}, {"label": "Steep setup", "summary": "Commenters say it is difficult to set up and manage, with a high learning curve.", "quotes": [{"id": 734445937, "item_id": 45902871, "url": "https://news.ycombinator.com/item?id=45902871", "by": "pat2man", "date": "2025-11-12", "stance": "recommend", "quote": "Kubernetes is quire resilient when set up properly, and the cost savings were significant.", "excerpt": "As someone who self hosted bare metal Kubernetes on my own rack, it's a lot of work to get it set up. We used RedHat Openshift which has a pretty good solution out of the box, but the learning curve was relatively high. That being said, once it was set up, there was not a lot of maintenance. Kubernetes is quire resilient when set up properly, and the cost savings were significant.", "context": "self-hosted Kubernetes on own rack", "use_case": "private cloud on bare metal", "switched_from": "", "story": "Kubernetes Is Your Private Cloud"}, {"id": 766030144, "item_id": 47876884, "url": "https://news.ycombinator.com/item?id=47876884", "by": "ianberdin", "date": "2026-04-23", "stance": "recommend", "quote": "k8s is insane peace of software. A right tool for a big problem.", "excerpt": "\u201cEverything is shit. Believe me. We will do something better, just believe me.\u201d Jokes aside: - k8s is insane peace of software. A right tool for a big problem. Not for your toys. Yes, it is crazy difficult to setup and manage. Then what? - cloud has bad and slow disk. BS. They have perfectly fast NVME. Something else? That\u2019s it. Why I am so confident? I used to setup and manage kubernetes for 2 years. I have some exp", "context": "managed Kubernetes for 2 years", "use_case": "large-scale orchestration", "switched_from": "", "story": "I Am Building a Cloud"}]}, {"label": "Resource overhead", "summary": "On many small nodes, monitoring and routing services consumed a large share of CPU.", "quotes": [{"id": 801847380, "item_id": 50115461, "url": "https://news.ycombinator.com/item?id=50115461", "by": "attilaolah", "date": "2026-10-16", "stance": "warn", "quote": "I realised I was burning a quarter of my fleet's CPU on just monitoring and routing services.", "excerpt": "I also used to run everything on K8s (hybrid Talos / Alpine+k0s on nodes too small for Talos). It is a fun exercise, but with many small nodes eventually I realised I was burning a quarter of my fleet's CPU on just monitoring and routing services. Eventually I switched to Systemd only, Clan-managed NixOS instances on bare metal. The killer feature for me is socket activation. It even works with Jellyfin, wrapped in s", "context": "self-hosts on many small nodes", "use_case": "homelab orchestration", "switched_from": "", "story": "What I self-host"}]}, {"label": "Poor sandbox fit", "summary": "Commenters found it a poor fit for rapidly churning sandboxes and agent environments needing local files or full desktops.", "quotes": [{"id": 783076352, "item_id": 48942272, "url": "https://news.ycombinator.com/item?id=48942272", "by": "no_circuit", "date": "2026-07-16", "stance": "warn", "quote": "Kubernetes likely isn't meant for rapid churn of workloads", "excerpt": "Sounds like the lesson learned is using the right tool for the job -- reusing Kubernetes in an existing cluster to spin up sandboxes is a fair initial path to start offering the service. But Kubernetes likely isn't meant for rapid churn of workloads, here sandboxes. The architecture to me seemed very similar to SeaweedFS [1] (Facebook Haystack [2]) except with an extra layer for sandbox-hosting nodes. Like requests g", "context": "", "use_case": "spinning up sandboxes", "switched_from": "", "story": "Scaling to 1M concurrent sandboxes in seconds"}, {"id": 802115296, "item_id": 50132206, "url": "https://news.ycombinator.com/item?id=50132206", "by": "everforward", "date": "2026-10-17", "stance": "warn", "quote": "I think k8s is a bad way to go for this", "excerpt": "I looked and will give it a try, but haven\u2019t had a chance yet because life. I had debated building something like this, and I really want to see what the other fork in the path looks like. One of the core tenants of mine is to not impose a workflow or set of tooling on the user. I specifically opted not to go with k8s because access to local files is going to be a pain, and I think people will still expect that. Ther", "context": "building their own agent container tool", "use_case": "sandboxing AI agents", "switched_from": "", "story": "OpenCode 2.0"}, {"id": 752472929, "item_id": 47029558, "url": "https://news.ycombinator.com/item?id=47029558", "by": "ed_mercer", "date": "2026-02-15", "stance": "warn", "quote": "VMs give you things containers struggle with", "excerpt": "For us, we actually moved away from k8s to dedicated VMs on Proxmox for our agents. We initially had a containerized environment manager running in k8s, but found that VMs give you things containers struggle with: full desktop environments with X11 for GUI automation, persistent state across sessions and dedicated resources per agent. Each agent gets their own Debian VM with a complete OS, which makes it much easier ", "context": "", "use_case": "running AI agents", "switched_from": "", "story": "Show HN: Kubernetes for AI Agents"}]}], "who": [{"by": "cwiggs", "context": "sysadmin/SRE/DevOps engineer", "url": "https://news.ycombinator.com/item?id=50146205", "date": "2026-10-18"}, {"by": "steezeburger", "context": "uses it at work too", "url": "https://news.ycombinator.com/item?id=50121892", "date": "2026-10-16"}, {"by": "cyberpunk", "context": "consultant working with clients", "url": "https://news.ycombinator.com/item?id=49937945", "date": "2026-10-02"}, {"by": "jabrotapu", "context": "freelancer building on-prem GPU clusters", "url": "https://news.ycombinator.com/item?id=49661006", "date": "2026-09-11"}, {"by": "sshine", "context": "homelab user", "url": "https://news.ycombinator.com/item?id=49219337", "date": "2026-08-07"}, {"by": "levl289", "context": "runs own mail server since early 2000s", "url": "https://news.ycombinator.com/item?id=49041845", "date": "2026-07-24"}, {"by": "hnlmorg", "context": "former K8s naysayer, experienced engineer", "url": "https://news.ycombinator.com/item?id=48669749", "date": "2026-06-24"}, {"by": "stevefan1999", "context": "building a platform to deploy open source apps on Kubernetes", "url": "https://news.ycombinator.com/item?id=48486140", "date": "2026-06-10"}, {"by": "Akuehne", "context": "self-hosts on a cheap mini PC", "url": "https://news.ycombinator.com/item?id=47974964", "date": "2026-05-01"}, {"by": "arjie", "context": "runs a homeserver", "url": "https://news.ycombinator.com/item?id=47971330", "date": "2026-04-30"}, {"by": "irusensei", "context": "uses it extensively at work", "url": "https://news.ycombinator.com/item?id=47887576", "date": "2026-04-24"}, {"by": "ianberdin", "context": "managed Kubernetes for 2 years", "url": "https://news.ycombinator.com/item?id=47876884", "date": "2026-04-23"}], "tags": ["for multi-node clusters", "for homelabs", "self-hostable on bare metal", "for GPU and AI workloads"], "use_cases": ["container orchestration", "deployment", "sandboxing ai agents", "running agents", "multi-machine home deployments", "managing self-hosted services"], "compared": [{"slug": "docker-compose", "name": "Docker Compose", "co": 8, "switched_from": 0, "switched_to": 1}, {"slug": "ansible", "name": "Ansible", "co": 5, "switched_from": 0, "switched_to": 2}, {"slug": "nixos", "name": "NixOS", "co": 4, "switched_from": 1, "switched_to": 1}, {"slug": "docker", "name": "Docker", "co": 7, "switched_from": 0, "switched_to": 0}, {"slug": "systemd", "name": "systemd", "co": 4, "switched_from": 0, "switched_to": 1}, {"slug": "proxmox", "name": "Proxmox", "co": 1, "switched_from": 0, "switched_to": 1}, {"slug": "nomad", "name": "Nomad", "co": 3, "switched_from": 0, "switched_to": 0}, {"slug": "helm", "name": "Helm", "co": 3, "switched_from": 0, "switched_to": 0}], "makers": [{"id": 796692978, "item_id": 49793311, "url": "https://news.ycombinator.com/item?id=49793311", "by": "rakyll", "date": "2026-09-21", "stance": "neutral", "quote": "I work on Kubernetes", "excerpt": "Disclaimer: I'm one of the co-creators of this project. AX is a layer that is closer to job orchestration, Agent Substrate. It's NOT an agentic framework. We use Antigravity for a few generative features but are abstracting away some of these components so anyone can bring their own. AX is trying to solve some tedious things everyone has to deal with. Layering execution with the underlying stateful worker, wiring up ", "context": "works on Kubernetes", "use_case": "", "switched_from": "", "story": "AX \u2013 Google\u2019s Open Agentic Orchestrator"}], "threads": [{"url": "https://news.ycombinator.com/item?id=50109715", "title": "What I self-host", "mentions": 5}, {"url": "https://news.ycombinator.com/item?id=47872324", "title": "I Am Building a Cloud", "mentions": 4}, {"url": "https://news.ycombinator.com/item?id=48665130", "title": "Zero-Downtime Deployments with Docker Compose \u2013 No Kubernetes Required", "mentions": 3}, {"url": "https://news.ycombinator.com/item?id=48872031", "title": "Rebuilding My Homelab with Compose, Ruby, IPv6, and No Kubernetes", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=48479452", "title": "Claude Desktop spins up a VM without no way of stopping it", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=47973939", "title": "Ask HN: How do you self-host your apps?", "mentions": 2}], "sentiment_by_month": {"2025-11": {"warn": 2, "recommend": 1, "compare": 1}, "2025-12": {"recommend": 1, "warn": 1, "neutral": 1}, "2026-01": {"recommend": 1, "neutral": 1, "warn": 1}, "2026-02": {"neutral": 1, "warn": 1, "recommend": 2}, "2026-03": {"neutral": 1}, "2026-04": {"neutral": 3, "recommend": 6}, "2026-05": {"neutral": 2}, "2026-06": {"neutral": 2, "warn": 3, "recommend": 2}, "2026-07": {"neutral": 3, "warn": 3, "recommend": 1}, "2026-08": {"recommend": 2}, "2026-09": {"neutral": 3, "warn": 1}, "2026-10": {"recommend": 3, "neutral": 3, "warn": 5}}, "quotes": [{"id": 804962705, "item_id": 50310169, "url": "https://news.ycombinator.com/item?id=50310169", "by": "fschuett", "date": "2026-10-30", "stance": "warn", "quote": "With a kubernetes cluster it's more or less based on hope and happy-path testing.", "excerpt": "It's not that crashes don't happen, it's just that they don't \"randomly\" happen. The error codes from bad hardware are predictable (against kernel bugs... well I use an \"allnoconfig\" kernel, which disables like 95% of the kernel). Bad hardware, bad network, bad RAM, all of these are simulateable in QEMU or Linux (tc netem / blkdebug). If I count my test suite for the ISO, I have ~2500 unit tests (placement groups tes", "context": "", "use_case": "", "switched_from": "", "story": "AI isn't replacing SREs, it's deskilling them"}, {"id": 803405488, "item_id": 50212843, "url": "https://news.ycombinator.com/item?id=50212843", "by": "doctorpangloss", "date": "2026-10-23", "stance": "warn", "quote": "\"we run everything on kubernetes, so deployment is a matter of using the least logical complement to it", "excerpt": "> We already run everything on Kubernetes, so deploying a service was mostly a matter of adding a Terraform module and a Helm chart. Haha \"we run everything on kubernetes, so deployment is a matter of using the least logical complement to it, and the lowest budget, flakiest and overbuilt alternative to plan manifests with none of the benefits of an operator\"", "context": "", "use_case": "", "switched_from": "", "story": "Can we replace a $150k Looker with a bucket of HTML files?"}, {"id": 803250848, "item_id": 50203178, "url": "https://news.ycombinator.com/item?id=50203178", "by": "dangoodmanUT", "date": "2026-10-23", "stance": "recommend", "quote": "believe me i love k8s, and i think agents wield it amazingly well", "excerpt": "schedulers that can handle spikier workloads generally believe me i love k8s, and i think agents wield it amazingly well, no more yaml-slop or kubectl (`k`) slop, just yell at claude to fix it", "context": "", "use_case": "deployment", "switched_from": "", "story": "OpenShell \u2013 A safe, private runtime for fleets of autonomous AI agents"}, {"id": 802339280, "item_id": 50146205, "url": "https://news.ycombinator.com/item?id=50146205", "by": "cwiggs", "date": "2026-10-18", "stance": "recommend", "quote": "for the past ~5 years I've been using Kubernetes and have helped companies migrate lots of workloads to k8s.", "excerpt": "This was a great read. I've been a sysadmin/SRE/Devops/Platform/infra ( Why are there so many different titles) engineers for many years, for the past ~5 years I've been using Kubernetes and have helped companies migrate lots of workloads to k8s. I've worked with many people who dislike k8s, but almost all of them had almost no experience with k8s. A lot of the anti-k8s people would eventually build out something sim", "context": "sysadmin/SRE/DevOps engineer", "use_case": "container orchestration", "switched_from": "", "story": "A skeptic's first contact with Kubernetes (2024)"}, {"id": 802138483, "item_id": 50133655, "url": "https://news.ycombinator.com/item?id=50133655", "by": "astralkraft", "date": "2026-10-17", "stance": "warn", "quote": "would like to be done with k8s altogether", "excerpt": "I think you're confusing this. Talos is kubernetes. It can only be as good or as bad as Kubernetes. This is different. A complete rethink of immutable infrastructure with a decade of additional hindsight. I switched from Talos to OKD. I honestly don't miss Talos, but would like to be done with k8s altogether.", "context": "", "use_case": "", "switched_from": "", "story": "Self-hosted infrastructure with Incus OS"}, {"id": 802115296, "item_id": 50132206, "url": "https://news.ycombinator.com/item?id=50132206", "by": "everforward", "date": "2026-10-17", "stance": "warn", "quote": "I think k8s is a bad way to go for this", "excerpt": "I looked and will give it a try, but haven\u2019t had a chance yet because life. I had debated building something like this, and I really want to see what the other fork in the path looks like. One of the core tenants of mine is to not impose a workflow or set of tooling on the user. I specifically opted not to go with k8s because access to local files is going to be a pain, and I think people will still expect that. Ther", "context": "building their own agent container tool", "use_case": "sandboxing AI agents", "switched_from": "", "story": "OpenCode 2.0"}, {"id": 801992834, "item_id": 50124552, "url": "https://news.ycombinator.com/item?id=50124552", "by": "verdverm", "date": "2026-10-16", "stance": "neutral", "quote": "I personally (1) run them in k8s", "excerpt": "We use GoModel for the gateway and visibility Sandboxing is less mature, but we use containers for our agents in CI. I personally (1) run them in k8s (2) replace my harness builtin tools with Dagger based implementation If an agent can delete a production database, that's PEBKAC imo. Our agents do not have access to such things. tl;dr - the solution is to shift-left to deterministic prevention, not stochastic detecti", "context": "", "use_case": "running agents", "switched_from": "", "story": "LLM says something, my tool track if it followed its promise on tools"}, {"id": 801950272, "item_id": 50121892, "url": "https://news.ycombinator.com/item?id=50121892", "by": "steezeburger", "date": "2026-10-16", "stance": "neutral", "quote": "I use it when I need to deploy a service across multiple machines.", "excerpt": "I'm sure you could achieve almost all the things I needed with docker compose, but I was using it at work and had designed some clean helm charts so just used them at home. I use it when I need to deploy a service across multiple machines. For personal projects that's pretty much my biggest decider. Otherwise I'm using the same docker compose setup I've been using for a while!", "context": "uses it at work too", "use_case": "multi-machine home deployments", "switched_from": "", "story": "What I self-host"}, {"id": 801913328, "item_id": 50119583, "url": "https://news.ycombinator.com/item?id=50119583", "by": "steezeburger", "date": "2026-10-16", "stance": "recommend", "quote": "you can't act like k8s and docker are overkill and fiddly simply because you haven't found a need for them yet.", "excerpt": "How is it a waste of time to health check your self hosted services? I use these services. I need them to work and be available. Bookmark manager, media server, my note taking app, a book server for friends, a few Discord bots I run, I've ran long running game servers for friends and a couple of little apps, home automation with CV ran on premise. We're obviously doing different things and have different priorities, ", "context": "professional software engineer", "use_case": "managing self-hosted services", "switched_from": "", "story": "What I self-host"}, {"id": 801864113, "item_id": 50116507, "url": "https://news.ycombinator.com/item?id=50116507", "by": "bengalister", "date": "2026-10-16", "stance": "neutral", "quote": "So definitely there is added complexity, especially on the iptables configuration.", "excerpt": "So far it is ok. I even have a usb zigbee dongle receiver in usb passthrough and the container in privileged mode. I am moving to the Apple ecosystem and I managed to have homekit bridge connected to Apple home. I had to use an avahi relay on the NixOS host for the discovery protocol to work. The home assistant is not the same network as my devices. The homekit bridge allows to configure the advertised IP. I use tail", "context": "", "use_case": "running Home Assistant", "switched_from": "", "story": "What I self-host"}, {"id": 801847380, "item_id": 50115461, "url": "https://news.ycombinator.com/item?id=50115461", "by": "attilaolah", "date": "2026-10-16", "stance": "warn", "quote": "I realised I was burning a quarter of my fleet's CPU on just monitoring and routing services.", "excerpt": "I also used to run everything on K8s (hybrid Talos / Alpine+k0s on nodes too small for Talos). It is a fun exercise, but with many small nodes eventually I realised I was burning a quarter of my fleet's CPU on just monitoring and routing services. Eventually I switched to Systemd only, Clan-managed NixOS instances on bare metal. The killer feature for me is socket activation. It even works with Jellyfin, wrapped in s", "context": "self-hosts on many small nodes", "use_case": "homelab orchestration", "switched_from": "", "story": "What I self-host"}, {"id": 801808257, "item_id": 50113016, "url": "https://news.ycombinator.com/item?id=50113016", "by": "steezeburger", "date": "2026-10-16", "stance": "neutral", "quote": "k8s is overkill for like 99% of my personal projects and like 75% of my professional projects.", "excerpt": "I like to have consistent architecture across all my services, and I like to see and configure everything in (mostly) one place. If you're running a single go binary, sure do whatever, but if you're running a go binary, and a *arr stack, and a bookmarking service, home assistant, frigate, etc, it gets out of hand pretty quickly. You trade a bit of upfront setup (I mean barely now in the age of LLMs) for easier automa", "context": "", "use_case": "large systems", "switched_from": "", "story": "What I self-host"}, {"id": 799007120, "item_id": 49937945, "url": "https://news.ycombinator.com/item?id=49937945", "by": "cyberpunk", "date": "2026-10-02", "stance": "recommend", "quote": "It's really not all that bad anymore.", "excerpt": "This was true maybe 8 years ago when all we had was buggy ceph drivers; but now the integration with the various cloud EBS-alikes is pretty good/seamless. VolumeSnapshots, WaitForFirstConsumer (so you're allocating in the right AZ as your STS comes up, etc etc). It's really not all that bad anymore. I'll take going to any client and having roughly the same storage setup anyday over the hell that was the mix of enterp", "context": "consultant working with clients", "use_case": "stateful workloads with cloud volumes", "switched_from": "", "story": "Pi 1.0"}, {"id": 796856370, "item_id": 49803523, "url": "https://news.ycombinator.com/item?id=49803523", "by": "purpleidea", "date": "2026-09-22", "stance": "warn", "quote": "Seeing Ansible+Kubernetes+K* stuff in their jobs offering does not inspire confidence", "excerpt": "I'd love to give someone at Verda a demo of https://github.com/purpleidea/mgmt/ to show them how they could actually do fast provisioning and automation. Seeing Ansible+Kubernetes+K* stuff in their jobs offering does not inspire confidence that they're innovating in building their core infra. You can't out-compete AWS or even the smaller clouds there. If you want to see a different perspective, lmk.", "context": "", "use_case": "infrastructure orchestration", "switched_from": "", "story": "Verda (Finland) raises $189M in Series B"}, {"id": 796770818, "item_id": 49798176, "url": "https://news.ycombinator.com/item?id=49798176", "by": "stevefan1999", "date": "2026-09-22", "stance": "neutral", "quote": "For me I went Docker and Kubernetes.", "excerpt": "Indeed. I still remember deploying GAE apps with Java WAR and Python Django as a kid 13 years ago. Then Heroku and now fly.io and Render. For me I went Docker and Kubernetes. Back then GAE has a really good free tier, so I used it to deploy a HTTP proxy to bypass school network", "context": "", "use_case": "deployment", "switched_from": "", "story": "Python Workers are now generally available"}, {"id": 794576096, "item_id": 49661006, "url": "https://news.ycombinator.com/item?id=49661006", "by": "jabrotapu", "date": "2026-09-11", "stance": "neutral", "quote": "Kubernetes across bare servers", "excerpt": "SEEKING WORK | India, remote | Part-time | Overlap with US, UK, EU hours I build on-prem GPU clusters and private AI apps that run entirely on the client's own hardware, so their data never leaves the building. About 10 years full-stack, the last few on GPU infrastructure. Currently running a live production cluster: 700-800 users sharing a small physical GPU pool. GPU-backed JupyterHub, single sign-on, per-user quot", "context": "freelancer building on-prem GPU clusters", "use_case": "GPU clusters on bare metal", "switched_from": "", "story": "Ask HN: Freelancer? Seeking freelancer? (September 2026)"}, {"id": 793898004, "item_id": 49618625, "url": "https://news.ycombinator.com/item?id=49618625", "by": "abejora", "date": "2026-09-08", "stance": "neutral", "quote": "We are now running a full kubernetes cluster on Talos Linux", "excerpt": "https://www.abejora.com/ - Timesheet registration software We try to differentiate in a few ways: * we try to be as fast and user friendly as possible, because nobody likes filling in timesheets: logical screens that align with the task you need to perform, keyboard support, flexible input... * democratic pricing. a lot of competitors hide the good stuff behind the expensive subscriptions, and per-seat pricing quickl", "context": "", "use_case": "hosting", "switched_from": "", "story": "Ask HN: Show your micro-SaaS"}, {"id": 787509392, "item_id": 49219337, "url": "https://news.ycombinator.com/item?id=49219337", "by": "sshine", "date": "2026-08-07", "stance": "recommend", "quote": "I moved my homelab to Kubernetes.", "excerpt": "I moved my homelab to Kubernetes. I just don\u2019t see the use of single VPS\u2019es. But I use the cluster as a training platform, so if I didn\u2019t need Kubernetes in other parts of my life, I probably would have stuck with my NixOS VPS\u2019es that all share config. I use kubenix to manage the cluster, of course. ;-)", "context": "homelab user", "use_case": "homelab cluster and training", "switched_from": "NixOS", "story": "The Nixpkgs core team has disbanded"}, {"id": 786739824, "item_id": 49171239, "url": "https://news.ycombinator.com/item?id=49171239", "by": "gerardnico", "date": "2026-08-04", "stance": "recommend", "quote": "You get also controlled rollout for free.", "excerpt": "You need to see Kubernetes as a wrapper of the init thread (systemd). If you fought with cgroup to tame resource (memory, cpu) consumption you know. You get also controlled rollout for free. Likewise, it\u2019s also a platform so you get installation receipts of third party software\u2019s for free. You install the Prometheus operator and, bam, you get declarative monitoring as code for free. The ticket entry is not free but o", "context": "", "use_case": "single-node deployment platform", "switched_from": "", "story": "Cheap Self-Hosted Kubernetes on Hetzner Cloud"}, {"id": 786024657, "item_id": 49126541, "url": "https://news.ycombinator.com/item?id=49126541", "by": "davkan", "date": "2026-07-31", "stance": "neutral", "quote": "a ceph backend on my k8s cluster", "excerpt": "I used to have a thunderbolt ring network for a ceph backend on my k8s cluster. On Linux load the thunderbolt_net kernel module and they show up the same as any other Ethernet interfaces. Just directly connect them and configure the network settings on each end using your normal tools. Latency was an order of magnitude higher than twisted pair but the throughput was greater than 10Gbps and I didn\u2019t have to spend on 1", "context": "", "use_case": "cluster", "switched_from": "", "story": "Getting 25 Gbps Thunderbolt Ethernet on My Mac Studio"}, {"id": 784669524, "item_id": 49041845, "url": "https://news.ycombinator.com/item?id=49041845", "by": "levl289", "date": "2026-07-24", "stance": "neutral", "quote": "now a single-node k8s instance with everything neatly managed", "excerpt": "I'd run my own email server since the early 2000's - in several forms over time. A VA Linux 2U rack server that I got from work (noisy, overkill), a personal Mac Mini (Frankenstein macport upgrade-resistant), and now a single-node k8s instance with everything neatly managed (an excuse to really dive into k8s from the ground up). I use dyndns to handle IP changes (incredibly rare with Frontier FIOS, actually). I use m", "context": "runs own mail server since early 2000s", "use_case": "running mail server", "switched_from": "", "story": "Self-host your mail server"}, {"id": 783076352, "item_id": 48942272, "url": "https://news.ycombinator.com/item?id=48942272", "by": "no_circuit", "date": "2026-07-16", "stance": "warn", "quote": "Kubernetes likely isn't meant for rapid churn of workloads", "excerpt": "Sounds like the lesson learned is using the right tool for the job -- reusing Kubernetes in an existing cluster to spin up sandboxes is a fair initial path to start offering the service. But Kubernetes likely isn't meant for rapid churn of workloads, here sandboxes. The architecture to me seemed very similar to SeaweedFS [1] (Facebook Haystack [2]) except with an extra layer for sandbox-hosting nodes. Like requests g", "context": "", "use_case": "spinning up sandboxes", "switched_from": "", "story": "Scaling to 1M concurrent sandboxes in seconds"}, {"id": 782934192, "item_id": 48933387, "url": "https://news.ycombinator.com/item?id=48933387", "by": "kjellsbells", "date": "2026-07-16", "stance": "warn", "quote": "you can hand massage exotic k8s setups on your fleet", "excerpt": "I sometimes feel that when the industry moved from pets to cattle, what really happened is that the cattle turned out to need an exotic farm to live on, negating the savings. You can have a few honking servers or you can hand massage exotic k8s setups on your fleet. Pick your poison, but dont delude yourself that the TCO of the latter is lower than the former.", "context": "", "use_case": "cluster orchestration", "switched_from": "", "story": "Making 768 servers look like 1"}, {"id": 782927088, "item_id": 48932943, "url": "https://news.ycombinator.com/item?id=48932943", "by": "maccard", "date": "2026-07-16", "stance": "recommend", "quote": "I think k8s is absolutely the right choice at that level.", "excerpt": "I disagree - I think k8s is absolutely the right choice at that level. We use ECS instead of k8s for small projects because the control plane is more than the actual service, but when you want container management, some sort of basic deployment management, service discovery, storage management and secret management, k8s is super simple to run and works pretty much out of the box. For a homelab, I'd expect you to want", "context": "", "use_case": "homelab container management", "switched_from": "", "story": "Rebuilding My Homelab with Compose, Ruby, IPv6, and No Kubernetes"}, {"id": 782894451, "item_id": 48930903, "url": "https://news.ycombinator.com/item?id=48930903", "by": "SMFloris", "date": "2026-07-15", "stance": "warn", "quote": "kubernetes is overkill for a small non-homogeneous home cluster.", "excerpt": "IMO, kubernetes is overkill for a small non-homogeneous home cluster. What I use and really recommend is using systemd +/- docker. It just becomes so darn simple. Do not go the compose route (that route is filled with sadness of the incomplete stacks because db container failed silently kind) - instead aim to decompose the compose files and write a separate systemd service file for each of them, you can then assign l", "context": "", "use_case": "homelab", "switched_from": "", "story": "Rebuilding My Homelab with Compose, Ruby, IPv6, and No Kubernetes"}, {"id": 782498129, "item_id": 48906133, "url": "https://news.ycombinator.com/item?id=48906133", "by": "afonsosoares", "date": "2026-07-14", "stance": "neutral", "quote": "On my case I use k8s and the backend pod use rust libsqld crate", "excerpt": "I could not agree more KISS is always the way to go. For those people that are interested you can have a sqld instance on a different VM with S3 backup. On my case I use k8s and the backend pod use rust libsqld crate with local first sql database file with remote sync to sqld. When pod start if db file is not available libsqld will try recover it from slqd otherwise it will just load local db file and sync.", "context": "", "use_case": "running backend pods", "switched_from": "", "story": "SQLite Is All You Need"}, {"id": 779865840, "item_id": 48741615, "url": "https://news.ycombinator.com/item?id=48741615", "by": "compumike", "date": "2026-06-30", "stance": "warn", "quote": "see if you're running \"Server Version: v1.36.[0,1,2]\"", "excerpt": "Author here! If you're running a Kubernetes cluster, I recommend you check `kubectl version` and see if you're running \"Server Version: v1.36.[0,1,2]\". If so, you may want to use the one-liner at the end of the article to check your \"process_resident_memory_bytes\" on each node, and consider restarting kubelet as a temporary workaround to tame the memory leak until v1.36.3 is released.", "context": "article author", "use_case": "", "switched_from": "", "story": "Fixing a kubelet Memory Leak in Kubernetes 1.36"}, {"id": 778769968, "item_id": 48673123, "url": "https://news.ycombinator.com/item?id=48673123", "by": "zug_zug", "date": "2026-06-25", "stance": "warn", "quote": "unnecessary complexity for almost all use cases", "excerpt": "I used to be critical of almost everyone who used it. Now that I've been learning the ecosystem for a couple years, I'm not as angry about it, and it doesn't affect me as much personally. I still think almost all of my complaints were valid: - not just container orchestration, more like a whole AWS solution (its own routing) - has single-points of failure that would be avoided if you just used aws - unnecessary compl", "context": "has been learning the ecosystem for a couple years", "use_case": "container orchestration", "switched_from": "", "story": "Zero-Downtime Deployments with Docker Compose \u2013 No Kubernetes Required"}, {"id": 778715984, "item_id": 48669749, "url": "https://news.ycombinator.com/item?id=48669749", "by": "hnlmorg", "date": "2026-06-24", "stance": "recommend", "quote": "The tooling has gotten better for setting up and managing K8.", "excerpt": "I\u2019m one of those top commenters who used to be a K8 naysayer. I\u2019m also definitely not young blood. The reason my position has changed is because: 1. The tooling has gotten better for setting up and managing K8. 2. In two of the last 3 jobs where we opted for a simplified alternative to k8, we came to regret that decision within a couple of years of that decision being made. If you\u2019re core architecture is changing on ", "context": "former K8s naysayer, experienced engineer", "use_case": "container orchestration", "switched_from": "", "story": "Zero-Downtime Deployments with Docker Compose \u2013 No Kubernetes Required"}, {"id": 778646208, "item_id": 48665388, "url": "https://news.ycombinator.com/item?id=48665388", "by": "nullpoint420", "date": "2026-06-24", "stance": "recommend", "quote": "I love how I can scale up all my services the same way, across clouds.", "excerpt": "100%. And a shared mental model. I love how I can scale up all my services the same way, across clouds. It's great.", "context": "", "use_case": "scaling services across clouds", "switched_from": "", "story": "Zero-Downtime Deployments with Docker Compose \u2013 No Kubernetes Required"}, {"id": 776897728, "item_id": 48556108, "url": "https://news.ycombinator.com/item?id=48556108", "by": "sshine", "date": "2026-06-16", "stance": "neutral", "quote": "I use AI for Kubernetes.", "excerpt": "I use AI for Kubernetes. On a day-to-day basis, it runs 80%+ of my kubectl commands. It\u2019s the most steroidal auto-complete I\u2019ve tried. But I do get dumber for it. What I do to compensate: - make it my duty to own every change, i.e. cognitively debt-free: - write summaries on every new thing I do (blog post, memo to colleague) - contribute documentation to the open-source projects I rely on - practice for CKA/CKAD cer", "context": "", "use_case": "cluster management", "switched_from": "", "story": "AskHN:How do you handle skill atrophy from using coding agents?"}, {"id": 776791760, "item_id": 48549485, "url": "https://news.ycombinator.com/item?id=48549485", "by": "mickael-kerjean", "date": "2026-06-15", "stance": "warn", "quote": "I used to run k8s on bare metal, troubleshooting something at least once a month", "excerpt": "I've had the opposite experience. I used to run k8s on bare metal, troubleshooting something at least once a month (DNS going down was a recurring favorite). The breaking point came with the churn in the ecosystem, got bitten by the deprecation of the community darling weave net cni plugin, the killing of the nginx ingress was the nail in the coffin knowing I had way too many annotation tight to the ingress that it w", "context": "ran k8s on bare metal", "use_case": "container orchestration", "switched_from": "Kubernetes", "story": "What job interviews taught me about Kubernetes"}, {"id": 775778240, "item_id": 48486140, "url": "https://news.ycombinator.com/item?id=48486140", "by": "stevefan1999", "date": "2026-06-10", "stance": "recommend", "quote": "I relied on so many features of what Kubernetes and its ecosystem provided", "excerpt": "> Ever hear of a large K8s migration going well ...well, AI infra stacks are more or less k8s/cloud native, especially on the inference side. Nvidia GPU operator plus KubeFlow makes deploying models easy, and I did that manually as well using just the ollama, llama.cpp containers and especially vllm operator. I'm not sure about what's the measuring notation of \"large\" is, but OpenAI had a blog post about how to manag", "context": "building a platform to deploy open source apps on Kubernetes", "use_case": "deploying open source apps and AI infra", "switched_from": "", "story": "Claude Desktop spins up a VM without no way of stopping it"}, {"id": 775755104, "item_id": 48484694, "url": "https://news.ycombinator.com/item?id=48484694", "by": "stevefan1999", "date": "2026-06-10", "stance": "neutral", "quote": "is still one of the mainstream container orchestrator if not the biggest one", "excerpt": "But Kubernetes, being a former Google project whilst donated to CNCF, is still one of the mainstream container orchestrator if not the biggest one (I believe we need to go over to supercomputers for that). Nomad is kind of dead and Apache Mesos is basically dead. Leaving Kubernetes as a natural moat, plus Google employees are still seemingly actively maintaining it. And also Go. While I'm not a Go guy (speaking as a ", "context": "", "use_case": "container orchestration", "switched_from": "", "story": "Claude Desktop spins up a VM without no way of stopping it"}, {"id": 768733057, "item_id": 48045816, "url": "https://news.ycombinator.com/item?id=48045816", "by": "tehlike", "date": "2026-05-06", "stance": "neutral", "quote": "then kubernetes managed the containe orchestration", "excerpt": "I used similar ipxe setup for robotic cluster - every robot booted from the same thing, then kubernetes managed the containe orchestration. it was fun.", "context": "", "use_case": "container orchestration", "switched_from": "", "story": "Diskless Linux boot using ZFS, iSCSI and PXE"}, {"id": 767963952, "item_id": 47997747, "url": "https://news.ycombinator.com/item?id=47997747", "by": "Akuehne", "date": "2026-05-03", "stance": "neutral", "quote": "I run k8s because everyone runs k8s now, and it's good to have a local environment to play with.", "excerpt": "Gitops. I use janky infrastructure as code to manage the services running on any stateful machines, and mainly just stuff I through together. I run k8s because everyone runs k8s now, and it's good to have a local environment to play with.", "context": "", "use_case": "self-hosting and learning", "switched_from": "", "story": "Ask HN: How do you self-host your apps?"}, {"id": 767599427, "item_id": 47974964, "url": "https://news.ycombinator.com/item?id=47974964", "by": "Akuehne", "date": "2026-05-01", "stance": "neutral", "quote": "All containers. Some just docker/podman, some one k8s cluster.", "excerpt": "Cheap mini pc on an isolated vlan. Running a cloudflared tunnel or reverse proxy to a vpc on digitalocean, maybe moving to Hetzner soon, or in addition to for failover/ha. All containers. Some just docker/podman, some one k8s cluster. Mainly it's just for fun. Except the cloud and local backups for our phones/Gcloud, which my wife will get really mad if it doesn't work.", "context": "self-hosts on a cheap mini PC", "use_case": "running self-hosted apps", "switched_from": "", "story": "Ask HN: How do you self-host your apps?"}, {"id": 767541280, "item_id": 47971330, "url": "https://news.ycombinator.com/item?id=47971330", "by": "arjie", "date": "2026-04-30", "stance": "neutral", "quote": "I use kube for most things so it's not a problem", "excerpt": "This is a valid concern, certainly. I use kube for most things so it's not a problem, but my homeserver and its apps run on quadlets that I manage. In my case, I just added a README.md in the server account folder that each project's CLAUDE.md or whatever is configured to read. Then it selects a port and sticks that in the document and to be honest I have a few tens of services and it works. Haha, a direct replacemen", "context": "runs a homeserver", "use_case": "running most services", "switched_from": "", "story": "I Got Sick of Remembering Port Numbers"}, {"id": 766201216, "item_id": 47887576, "url": "https://news.ycombinator.com/item?id=47887576", "by": "irusensei", "date": "2026-04-24", "stance": "recommend", "quote": "it was not that bad as people says it is", "excerpt": "I use kubernetes extensively at work. I don't manage the kubernetes cluster anymore since now we have a team that runs centralized services and you can request a namespace with a quota. But back when my team had a dedicated Azure Kubernetes cluster it was not that bad as people says it is and the biggest hassle was the extremely short lived support for each version. Then I started to realize most people who complain ", "context": "uses it extensively at work", "use_case": "container orchestration at work", "switched_from": "", "story": "I Am Building a Cloud"}, {"id": 766030144, "item_id": 47876884, "url": "https://news.ycombinator.com/item?id=47876884", "by": "ianberdin", "date": "2026-04-23", "stance": "recommend", "quote": "k8s is insane peace of software. A right tool for a big problem.", "excerpt": "\u201cEverything is shit. Believe me. We will do something better, just believe me.\u201d Jokes aside: - k8s is insane peace of software. A right tool for a big problem. Not for your toys. Yes, it is crazy difficult to setup and manage. Then what? - cloud has bad and slow disk. BS. They have perfectly fast NVME. Something else? That\u2019s it. Why I am so confident? I used to setup and manage kubernetes for 2 years. I have some exp", "context": "managed Kubernetes for 2 years", "use_case": "large-scale orchestration", "switched_from": "", "story": "I Am Building a Cloud"}, {"id": 765063552, "item_id": 47816472, "url": "https://news.ycombinator.com/item?id=47816472", "by": "adamcharnock", "date": "2026-04-18", "stance": "recommend", "quote": "using Kubernetes to distribute workloads across servers and provide HA", "excerpt": "This is something we've[0] done a number of times for customers coming from various cloud providers. In our case we move customers onto a multi-server (sometimes multi-AZ) deployment in Hetzner, using Kubernetes to distribute workloads across servers and provide HA. Kubernetes is likely a lot for a single node deployment such as the OP, but it makes a lot more sense as soon as multiple nodes are involved. For backups", "context": "runs a hosting/infra company moving customers to Hetzner", "use_case": "HA on Hetzner multi-server", "switched_from": "", "story": "Migrating from DigitalOcean to Hetzner: From $1,432 to $233 With Zero Downtime"}, {"id": 752628816, "item_id": 47039301, "url": "https://news.ycombinator.com/item?id=47039301", "by": "himata4113", "date": "2026-02-16", "stance": "recommend", "quote": "I have been running 5 k8s-flavored clusters for more than half a decade between 1 to 40 nodes.", "excerpt": "Kubernetes solves real problems for the 1% who need it. The other 99% are paying a massive complexity tax for capabilities they never use, while 87% of their provisioned CPU sits idle. is where the author is just wrong: - abstracts away ssh - makes it pretty unnecessary - rbac multi tenancy - better automations - orchestating more than one cluster - better infra as code - provisions are as good as you make them, if y", "context": "runs 5 clusters for over half a decade", "use_case": "container orchestration", "switched_from": "", "story": "Docker Swarm vs. Kubernetes in 2026"}, {"id": 747781185, "item_id": 46736324, "url": "https://news.ycombinator.com/item?id=46736324", "by": "jesse_dot_id", "date": "2026-01-23", "stance": "warn", "quote": "K8s is extreme overkill for a high percentage of most of the shops who are using it", "excerpt": "We use swarm in production and love it. K8s is extreme overkill for a high percentage of most of the shops who are using it, in my estimation.", "context": "", "use_case": "container orchestration", "switched_from": "", "story": "What Has Docker Become?"}, {"id": 735137456, "item_id": 45946091, "url": "https://news.ycombinator.com/item?id=45946091", "by": "silasb", "date": "2025-11-16", "stance": "compare", "quote": "k8s is immensely powerful but often feels like overkill for teams that just need simple orchestration", "excerpt": "I'm not trying to take a shot at the OP, but I keep seeing posts labeled \"Production-Grade\" that still look more like pet systems than cattle. I'm struggling to understand how something like this can be reproduced consistently across environments. How would you package this inside a Git repo? Can it be managed through GitOps? And if we're calling something production-grade, high availability should be a baseline requ", "context": "", "use_case": "container orchestration", "switched_from": "", "story": "Production-Grade Container Deployment with Podman Quadlets \u2013 Larvitz Blog"}, {"id": 735125713, "item_id": 45945357, "url": "https://news.ycombinator.com/item?id=45945357", "by": "rootnod3", "date": "2025-11-16", "stance": "warn", "quote": "I used to love both, Kubernetes and Nix. But after a few years of using both I felt like the abstraction levels are a bit too deep.", "excerpt": "I used to love both, Kubernetes and Nix. But after a few years of using both I felt like the abstraction levels are a bit too deep. Sure, it's easy to stand up a mail server in NixOS, or to just use docker/kubernetes to deploy stuff. But after a few years it felt like I don't have a single understanding of the stack. When shit hits the fan, it makes it very difficult to troubleshoot. I am now back on running my serve", "context": "runs own servers", "use_case": "deploying services", "switched_from": "", "story": "Run Nix Based Environments in Kubernetes"}, {"id": 734445937, "item_id": 45902871, "url": "https://news.ycombinator.com/item?id=45902871", "by": "pat2man", "date": "2025-11-12", "stance": "recommend", "quote": "Kubernetes is quire resilient when set up properly, and the cost savings were significant.", "excerpt": "As someone who self hosted bare metal Kubernetes on my own rack, it's a lot of work to get it set up. We used RedHat Openshift which has a pretty good solution out of the box, but the learning curve was relatively high. That being said, once it was set up, there was not a lot of maintenance. Kubernetes is quire resilient when set up properly, and the cost savings were significant.", "context": "self-hosted Kubernetes on own rack", "use_case": "private cloud on bare metal", "switched_from": "", "story": "Kubernetes Is Your Private Cloud"}]}