I spent probably a weeks worth of hours learning more #kubernetes so I could save $60 a month.
I have a nice 3 node kube cluster with a 2 node #keepalived #haproxy TCP load balancer. All on #ARM VPS.
Haproxy ingress
#ExternalDNS operator
#CertManager
#RookCeph
#ArgoCD
#KeyCloak
#ValKey
#Mastodon
#CloudNativePG #Postgresql
Kubernetes 1.33 stabilisiert 24 Features und bringt zahlreiche Neuerungen
Gespickt mit neuen Alpha-, Beta- und stabilen Features soll die Container-Orchestrierung Entwicklern mehr Kontrolle bieten und Latenzen verringern.
Kubernetes 1.33: упорядоченное удаление ресурсов, изменение алгоритма CrashLoopBackOff и декларативная валидация
Сегодня официально выпустили очередную версию Kubernetes — 1.33. Собрали все 64 изменения в одном материале. Из основных нововведений: упорядоченное удаление ресурсов в пространстве имён на основе логических зависимостей и соображений безопасности, декларативная валидация для нативных API-типов, расширение механизма CredentialProvider, доступ подов к информации о топологии кластера, изменение алгоритма выдержки CrashLoopBackOff, обязательная аутентификация при извлечении private-образов из репозиториев и многое другое.
Been doing some Kubernetes API design today. It's thoughtful but not in quite the way other software design has felt.
I suppose the constraints make it easier to avoid some common mistakes. Spending quite a bit of effort though just wrangling GoLang types and kubebuilder annotations to try and get it how it ought to be.
Harness is hiring Implementation Engineer
#python #aws #azure #cicd #gcp #kubernetes #terraform
Bengaluru, India
Full-time
Harness
Job details https://jobsfordevelopers.com/jobs/implementation-engineer-at-harness-io-dec-5-2024-919748?utm_source=mastodon.world&utm_medium=social&utm_campaign=posting
#jobalert #jobsearch #hiring
As of tonight I've moved substantially all of my #HomeLab workloads into my #Kubernetes cluster. Tonight's big push was the Minecraft server. Not doing anything fancy, just a simple StatefulSet, a pvc, and a MetalLB service.
There are a few lingering issues. I need to figure out Multus so #HomeAssistant can get multicast back. I'm considering moving all of my zigbee/zwave/etc gateways into the cluster and running a simple ser2net docker container on the edge nodes.
I also want to move VMSave. It's running on a somewhat expensive DigitalOcean VM today and would be a tiny smidge on the cluster. It's a bit more complicated because it's a super old Rails app currently deployed with Dokku, so it'll keep for another day.
Marqeta is hiring Senior Software Engineer - CPM Disputes
#java #react #springboot #aws #azure #cicd #gcp #kubernetes #mysql #sql #seniorengineer Vancouver, Canada Full-time Marqeta
Job details https://jobsfordevelopers.com/jobs/senior-software-engineer-cpm-disputes-at-marqeta-com-mar-12-2025-2357e9?utm_source=mastodon.world&utm_medium=social&utm_campaign=posting
#jobalert #jobsearch #hiring
This week's "fun" project: getting #Clickhouse running on #kubernetes at home, with data sharding and redundancy.
I've had a few speedbumps, including dirt in the optics on a redundant network link and an amazingly dumb MTU problem, but it seems to be working finally. I'm now doing a bit of testing to better understand how self-hosted Clickhouse does clusters.
It looks like *if you use their cloud product*, then data sharding and replication Just Works, but if you set up a cluster yourself then you need to declare everything up front when creating tables. So instead of creating a single table via `create table foo (...) ENGINE = MergeTree`, you need to do `create table foo_shard on cluster X (...) ENGINE = ReplicatedMergeTree(...)` in order to create replicated sub-tables per shard, and then add `create table foo on cluster X as foo_shard ENGINE=Distributed(...)` over the top of the per-instance ReplicatedMergeTree shards.
It's (mostly?) just a DDL thing, so querying seems to work as expected, but it's *strange* to create a cluster with a defined shard size and replication level, and then have to repeat yourself per-table in order to use them as declared.
Even better, this all "works" if you declare the per-shard tables as `MergeTree` instead of `ReplicatedMergeTree`, except your data isn't replicated. I watched the disk usage climb on 1/3rd of my nodes while the other 2/3rds sat idle, and had to go back and re-read docs. That's *particularly* surprising, as it could easily lead to data loss.
I'm doing yet another test copy of 1.2B rows of log data (!) right now, and then I'll start looking at what I need to do to cut over.
Also, I'll probably need to do some querying to see why I have 1.2B rows of log data and where it came from. That seems excessive for a couple weeks of logging at home.
After many days of trial and error finally I was able to switch from Docker’s docker-in-docker image to #Podman to run #RustLang integration tests using #Testcontainers in #WoodpeckerCI. It was quite a journey considering my setup, as the agent is deployed into #kubernetes. It definitely deserves a longer post.
(vdx.hu s/ybIm)
Week ending April 20: 1.33 released! Plus blogs, shoutouts, and patch releases.
Kubernetes Pods Are Inheriting Too Many Permissions – Source: www.darkreading.com https://ciso2ciso.com/kubernetes-pods-are-inheriting-too-many-permissions-source-www-darkreading-com/ #rssfeedpostgeneratorecho #DarkReadingSecurity #CyberSecurityNews #DARKReading #Kubernetes
Kubernetes v1.33: Octarine - https://kubernetes.io/blog/2025/04/23/kubernetes-v1-33-release/ #Kubernetes
Stop manual edge deployments! 
Our free webinar with Penguin Solutions on May 6 reveals how VMs, containers, and #Kubernetes enable automated, zero-touch deployments and streamlined lifecycle management. Save your spot 
https://bit.ly/PenguinSolutionsWebinar
New Kubernetes Release
Kubernetes v1.33.0
https://github.com/kubernetes/kubernetes/releases/tag/v1.33.0
Недостатки Istio по сравнению с Cilium: подробное объяснение
В этой статье мы разберём основные недостатки Istio в сравнении с Cilium Service Mesh, чтобы даже начинающий разработчик мог понять, в чём разница и почему некоторые команды выбирают Cilium вместо Istio.
If you're a fan of #TalosLinux, we've now added the ISO of v1.9.5 to our pool of available operating systems on #Hetzner Cloud servers. Changelog here:
https://docs.hetzner.cloud/changelog#2025-04-23-talos-linux-v195-iso-now-available
The upstream docs have been updated as well:
https://www.talos.dev/v1.9/talos-guides/install/cloud-platforms/hetzner/
@mshiltonj Umm, no. From 2024-08-13 to 2025-10-28 is a little over 14 months. Note the year.
Still fast, but not *that* fast.
Wow. A #k8s version goes from "release" to "end of life" in LESS THAN THREE MONTHS. What an upgrade and maintenance treadmill. #devops #kubernetes.
EDIT: FALSE ALARM. I WAS WRONG. It's a year and three months.
День из жизни облачной команды: как мы провели Demo Day
Привет, Хабр! В этом материале мы не расскажем о рабочих кейсах, технических решениях и привычных темах для нашего блога. Лучше — мы покажем немного всего этого и поделимся впечатлениями о первом Demo Day облачной команды Рег.ру. Внутри: краткое ревью докладов, анонсы новых облачных релизов, которые мы планируем к запуску, и немного фото. Полетели!
arc) 
