I created an Open Source tool to fork Kubernetes environments it is like "Git Fork" but for k8s.

Hi Folks,

I created an open-source tool that lets you create, fork, and hibernate entire Kubernetes environments.

With Forkspacer, you can fork your deployments while also migrating your data.. not just the manifests, but the entire data plane as well. We support different modes of forking: by default, every fork spins up a managed, dedicated virtual cluster, but you can also point the destination of your fork to a self-managed cluster. You can even set up multi-cloud environments and fork an environment from one provider (e.g., AWS) to another (e.g., GKE, AKE, or on-prem).

You can clone full setups, test changes in isolation, and automatically hibernate idle workspaces to save resources all declaratively, with GitOps-style reproducibility.

It’s especially useful for spinning up dev, test, pre-prod, and prod environments, and for teams where each developer needs a personal, forked environment from a shared baseline.

License is Apace 2.0 and it is written in Go using Kubebuilder SDK

***https://github.com/forkspacer/forkspacer*** \- source code

Please give it a try let me know, thank you

https://redd.it/1olo2l4
@r_devops

Understanding Docker Multi-platform Builds with QEMU

https://cefboud.com/posts/qemu-virtualzation-docker-multi-build/


https://redd.it/1olod4p
@r_devops

VS Code extension for dependency CVE scanning

VulScan-MCP scans project manifests for security vulnerabilities.



Queries NVD and OSV APIs for CVE data. Integrates with GitHub Copilot via Model Context Protocol.



Supports npm, pip, Maven, Go modules, Cargo, and more.



Open source: https://github.com/abhishekrai43/VulScan-MCP



Try it if you want CVE scanning in your editor.

https://redd.it/1olpe9t
@r_devops

A simple shell noscript that creates rootless podman containers to automate any task, building of github projects, kernels, applications etc

Denoscription: A simple shell noscript that uses buildah to create customized OCI/docker images and podman to deploy rootless containers designed to automate compilation/building of github projects, applications and kernels, including any other conainerized task or service. Pre-defined environment variables, various command options, native integration of all containers with apt-cacher-ng, live log monitoring with neovim and the use of tmux to consolidate container access, ensures maximum flexibility and efficiency during container use.

Url: https://github.com/tabletseeker/pod-buildah

https://redd.it/1olzbox
@r_devops

A simple shell noscript that creates rootless podman containers to automate any task, building of github projects, kernels, applications etc.

Denoscription: A simple shell noscript that uses buildah to create customized OCI/docker images and podman to deploy rootless containers designed to automate compilation/building of github projects, applications and kernels, including any other conainerized task or service. Pre-defined environment variables, various command options, native integration of all containers with apt-cacher-ng, live log monitoring with neovim and the use of tmux to consolidate container access, ensures maximum flexibility and efficiency during container use.

Url: https://github.com/tabletseeker/pod-buildah

https://redd.it/1olz4ug
@r_devops

Exploring low latency audio AI agents for live communication 🎧

I’ve been messing with some real-time audio based AI Agents to handle latency, reasoning, and synchronization when assisting during live human interviews, meetings and conferences etc.

The best examples I’ve found so far are Cogniear, LockedIn and Parakeet AI agents, all focused on real-time live spoken coaches rather than text.

-Cogniear.com works as an end-to-end reasoning loop: listens to and understands to whisper a full, spoken response in under 2 seconds.

-LockedInAI acts as a contextual tone coach, analyzing your confidence and phrasing during meetings.

-ParakeetAI focuses on improving clarity, cadence, and emotional delivery in real time.

It feels like early-stage “symbiotic audio reasoning” where human speech and AI processing overlap instead of alternating turns.

Questions for devs:

\-What’s the most efficient way to reduce inference lag in real-time voice reasoning systems?

\-How can multi-agent voice models maintain coherent dialogue flow without desyncing?

\-Anyone try prototyping something similar using streaming inference or hybrid STT/TTS pipelines?

Has anyone here tried something like   that?Would love to hear your experiences with any real-time audio based AI Agents

https://redd.it/1om2cxh
@r_devops

Hosting my CI/CD setup on a smaller EU cloud turned out smoother than I expected

I’ve been testing a few European clouds for my CI/CD setup, mainly Xelon.ch, Hetzner, and Scaleway. All of them did fine tbh , but I really liked how smooth and minimal the Xelon dashboard feels.

Setup took maybe 15 mins total, Jenkins + containers + auto backups. Everything runs from ISO-certified Swiss data centers, which adds a nice layer of trust.

Hetzner’s pricing is great, Scaleway’s UI is clean, but Xelon felt like the right mix of speed, compliance, and stability.

Anyone else here using EU-based clouds for pipelines or container work?

https://redd.it/1om5tqo
@r_devops

GraphQL Batching Attacks: How 100 Queries Become 10,000 Database Calls 📊

https://instatunnel.my/blog/graphql-batching-attacks-how-100-queries-become-10000-database-calls

https://redd.it/1om7jq7
@r_devops

Feeling stuck in DevOps tutorial hell for 5+ years — need guidance, structure, mentor, or cohort. How do I escape this cycle and make the switch?

Hi everyone. I’m a Senior Software Developer in Test (SDET) from India. For years I’ve been trying to transition into DevOps/SRE… but I feel completely stuck and lost.

My background:
I’ve been working professionally with Selenium, Maven, Jenkins, GitHub Actions, and automation frameworks. I also have some scattered hands-on touch-points with Docker, Kubernetes, Terraform, Ansible, Linux, Cloud… but NOTHING fully end-to-end production level. Only small experiments, tutorial-based setups and minor infra work for automation.

For the past 5-6 years, I’ve been trying to learn DevOps solo — watching endless Udemy courses, YouTube channels, reading various books, taking notes, doing bits and pieces here and there… but there is NO real direction or structure. It feels like I know a little of EVERYTHING, but I’m not DEEP in anything. I’m basically a “Jack of all tools, master of none.”

# The real problem:

DevOps is extremely broad.
Looking at AWS alone feels like a 2 year study.
Linux itself could take 1 year deeply.
Kubernetes is practically its own universe.
Every roadmap online looks endless — like a 10 year journey.

So what happens is:

I jump tool → to tool → to tool → to resource → to another course
without ever completing a structured path.

This has led me into a **never ending tutorial hell** for YEARS.

And this is starting to affect me mentally/emotionally.
I feel depressed because I do so much effort, consume so much content, but I still don’t feel confident enough to call myself a real DevOps engineer.

# What I need:

I don’t want another random list of videos/courses to watch.

I need:

* STRUCTURE
* ACTIONABLE sequence
* A clearly defined set of sub-skills
* EXACT things to learn in each major area (Linux → Docker → K8s → IaC → Cloud → CI/CD etc.)
* REAL capstone projects end to end that simulate real production DevOps architecture
* Guidance on how to network / get referrals / find DevOps jobs in this AI dominated environment

Example of what I mean by direction:

* “Here is the exact problem statement.”
* “Design this workload on AWS with these components.”
* “Configure DNS this way.”
* “Implement load balancers like this.”
* “Use Ansible here.”
* “Deploy this app with Kubernetes here.”
* “Document it into a portfolio.”
* “Do 3-4 such major capstones — that is enough to confidently apply for Senior DevOps roles.”

This is the kind of clarity I am desperately missing.

# What I’m searching for now:

* Someone who has successfully transitioned — and can mentor me (even paid mentor is fine)
* Or a cohort / group of people preparing for DevOps roles together
* Or a structured learning community with consistency and direction
* Or experienced DevOps engineers who can tell me the minimum essential path (without drowning me in infinite tool lists)

I’m NOT asking for hand-holding where someone does everything for me.

I just need a guiding force who says:

* “Do THIS next.”
* “Focus on THIS area.”
* “Complete THIS project.”

I can work extremely hard if I know I’m working in the right direction.

Right now I feel like I’m digging myself deeper into knowledge without outcomes. It feels like a hole that I cannot climb out of alone.

If anyone here has gone through this transition:
How did you break out?
How did you find the right direction?
How did you filter out noise vs essentials?
Where did you find the right mentor/community/cohort?

Any guidance here would genuinely help me get unstuck.

https://redd.it/1om8kb1
@r_devops

How do you track your cloud spend? Per instance daily, or monthly totals across all servers?

Hey folks,
I’m curious how other teams handle cloud cost tracking and reconciliation in day-to-day operations.

In our setup, we run about 10 instances with mixed workloads (compute, storage, and network). I’m wondering how you usually keep an eye on costs. Do you track daily usage per instance like CPU hours, storage, and bandwidth? Or do you mostly review monthly totals across all servers?
What’s been your best practice for keeping visibility without spending half your week digging through usage reports?

https://redd.it/1omacjj
@r_devops

What tech stack or setup do you use that gives you similar capabilities to a full-featured PaaS?

I’ve been comparing hosting options and noticed that services like Linode or DigitalOcean, ... don’t really offer much in terms of DevOps automation or collaboration tools. Some PaaS platforms, on the other hand, provide pretty advanced features, like full, application-aware cluster snapshots (flushing MySQL/Redis/Solr before taking them), instant Copy-on-Write environment clones per Git branch, and seamless Git-based deployments.

You can debug live environments, integrate easily with GitHub/GitLab/Bitbucket, and even host multiple apps (frontends, WordPress, microservices, etc.) within a single project. It’s incredibly convenient for team-based development, though obviously, it’s not cheap.

I know it’s difficult to fully replicate what modern PaaS platforms offer with, but I’d love to know what kind of tech stack and methodologies people are using to get close.

I’m not a DevOps engineer, just a developer who wants to experiment with this kind of setup for PHP CMS projects like WordPress and Drupal, mostly for learning and training purposes and personal projects.

https://redd.it/1ombw30
@r_devops

Lost in the journey

I'm working as a programmer since 1 year and a half, but lately the more i try to understand the more i get confused by the load of things there are and i question myself "why all of these? How can i improve knowing i'll never use these things on my own projects?".

In this 1,5 year i worked in two companies:
-one used old school programming: html+css+js+php all in the same file, no versioning, programming in production, no IDE and the client was at european level
-the second was hyper modern: python django+vue+hg+ide+ci/cd+super abstraction+proprietary models+docker+staging/prod and different servers


The first one was hard because it was difficult to find what to do and where, lost in 3/4k rows of files with everything mixed together.

But the second one is even harder because the abstraction level is so high that there is a model that does what you must do, but it's hidden somewhere in a combination of hundreds of imports and files everywhere and if you don't know these proprietary models you'll never understand what they do.

And this means zero creativity, everything is so abstract that even the smallest fix requires many steps of integration and you may miss something in the process..

So i'm here spending hours or even days to try to understand the flow, knowing that outside the work i cannot study these things and while i'm at work these things may be upgraded.. so everytime i program i feel like i'm moving super slowly, even the smallest fix requires hours and hours and without the certainity to do that right..

What should i do?
Thanks

https://redd.it/1omc654
@r_devops

How do you think working in ops has changed you as a person?

I am pondering this question myself and have no firm ideas yet, and thought the community might find it an interesting question

https://redd.it/1omdt9f
@r_devops

Need Advice !

Hi Folks,

Please take 2–3 minutes to read this — your advice would be truly appreciated.

I’m a 26-year-old professional seeking guidance. Please find my background below:

Experience: 3.9 years (MNC)
Certifications: 3x AWS
Skills: Kubernetes, Terraform, AWS, GitHub Actions, EKS, Docker, CI/CD

What I do in my Homelab:
I regularly practice deploying Flask applications on Docker and EKS containers, create Terraform modules, build GitHub Actions workflows, and work on Python automation projects. I also develop Terraform and EKS projects in my free time.

What I do in my current organization:

1. Handling repetitive ServiceNow tickets
2. Server patching (simple 2–3 step process)
3. Performing vulnerability remediation (manually installing updated software like 7-Zip, Notepad, etc.)
4. No exposure to Terraform, EKS management, or major incident handling (P1/P2). I’m in a comfort zone that doesn’t challenge me or contribute to my growth.


Looking for Devops Opportunities

I’m considering resigning from my current organization without having another offer in hand, as the current work environment feels stagnant and offers minimal learning opportunities.

From your perspective, would it be wise to take this step now? I’d appreciate your honest opinions and suggestions.

My financial situation is good 👍, but the only thing holding me back is the fear of not finding a job after resigning.

https://redd.it/1omfntj
@r_devops

Good source for DevOps fundamentals and terms?

Hello everyone,

I got a job as Machine Learning Engineer but have a background in Mechatronics/ Robotics. I did my practical thesis in ML development for industrial implementation.

Therefore I know how to build and train ML models, but I am not an software engineer.

Does someone have good resources for me? Or good roadmap to learn software engineering/devops fundamentals and terminology?
By the way I like structured sources 👌🏽

https://redd.it/1omf0ax
@r_devops

Hey guys need guidance

Hey guys I am preparing for switch from my first company
Some background, after college I got offer in as cloud ops engineer been working in same company for almost 2.5 years now thinking of switching I mainly have 3 questions
1. Is market favourable for the switch as cloud or DevOps enginey
2. As per my experience of 2.5 years how much salary hike I can expect current in hand is 6
3. I got experience in aws gcp somewhat in k8s, also know linux was from coding background so know basic in programming as well so anything you suggest I should run and polish my skillset
4. If you could give me some projects that could help in strengthening the resume , like general idea will be good aswell thanks in advance

https://redd.it/1omjtbq
@r_devops

Tomorrow my first day as devops engineer, any tips? Anything would be appreciated. Bit anxious tbh

I have been on rest for like 5 months due to acl injury and tomorrow is the first day as a devops engineer (intern for the first three months tho). My first job. Wooow excited tbh. Actually doesn't have much experience in this role or field, was into cybersecurity before. Any tips or suggestions would be appreciated.

https://redd.it/1omldja
@r_devops

When a missing flag breaks your deploy: -D vs -P in Java builds

I once hit a weird deployment issue because I confused -Denv=prod with -Pprod. Wrote a short note to help newer devs understand what actually happens under the hood.

It’s aimed at junior engineers working on CI/CD or build noscripts who want to know when to use which flag.

Read it here -> https://medium.com/stackademic/two-tiny-flags-that-confuses-java-devs-d-and-p-in-java-and-maven-5dfd0e04455f?sk=6b0d660c1a031576b629d7979054fd88

https://redd.it/1omn1hr
@r_devops

Looking for advice - I've built an AI-augmented Network Configuration and Troubleshooting Agent - worth it?

While it may look like self-promo, I'm looking for a feedback from fellow network engineers who had hands-on experience with AI agents and their implementations.

To provide more context:

As we all know, network devices (routers, switches, firewalls) are configured via CLI over SSH, sometimes REST/API. All traditional automation (Ansible, Python noscripts) requires predefined playbooks for every scenario. I wanted something that could:

* Reason about network problems dynamically
* Consult vendor documentation before acting
* Handle multi-vendor environments without rigid playbooks
* Operate safely with strong guardrails, lots of strong guardrails
* Work in a multi-tenant architecture

**Key parts:**

**RAG Implementation**

* AWS OpenSearch cluster with vendor documentation (Cisco, Juniper, Fortinet, etc.)
* Chunking strategy: per-command documentation + contextual sections
* Metadata tagging: device type, OS version, command category
* Retrieval: hybrid search (semantic + keyword) to find relevant docs before execution
* Challenge: Vendor docs are inconsistent in format/quality - had to build custom parsers per vendor

**Tool Design**

* `ssh_execute`: Run commands with device context awareness
* `get_device_config`: Retrieve current configs for analysis
* `consult_docs`: RAG retrieval before any config change
* `validate_syntax`: Pre-check commands against vendor syntax rules
* `rollback`: Automatic config snapshots before changes

**Guardrails**

* Restricted command whitelist/blacklist per environment
* Read-only mode by default
* Required approval workflow for config changes
* Device type validation (won't run Cisco commands on Juniper)
* Rate limiting on CLI execution
* Automatic rollback on detected errors

**Multi-Agent Pattern (Considering)** Currently single-agent with tool use, but exploring:

* Planner agent: decides approach
* Execution agent: runs commands
* Validation agent: checks results
* Documentation agent: pure RAG queries

Not sure if the added complexity is worth it yet.

Here is a snippet of how it replies when asked about configuring ZTNA server on the firewall device:
[https://imgur.com/a/dUjQrV3](https://imgur.com/a/dUjQrV3)
[https://imgur.com/a/fdIgr91](https://imgur.com/a/fdIgr91)

It first queries the devices, then searches through the docs for the info:
[https://imgur.com/a/PTqzTnN](https://imgur.com/a/PTqzTnN)

I picked two random products just to see how it responds when it comes do maintenance window recommendations.
[https://imgur.com/a/qbMpDfa](https://imgur.com/a/qbMpDfa)
[https://imgur.com/a/oPuhg1o](https://imgur.com/a/oPuhg1o)

**Where I would love your feedback:**

1. Which vendor tasks are the biggest time sinks: SR creation, RMA, firmware advisories, license renewals, config drift, SLA tracking, something else?
2. If you’ve used agents, where did they help/hurt (triage, enrichment, execution, hallucinations, RBAC/approvals)?
3. Integration realities: ConnectWise/Autotask, common RMMs/ITSMs, data residency, SSO, on-prem constraints.
4. What metrics would convince you this is worth it (MTTA/MTTR, SLA hit rate, case duration, renewal touch time, engineer hours saved)?
5. Any absolute non-starters (lock-in, privacy, vendor T&Cs, API rate limits)?

Not a pitch — trying to be realistic about this thing. When we were building it - things like compliance and scalability were first in mind.

https://redd.it/1ommjxd
@r_devops

I made an Android app to manage my Docker containers on the go

Hello Everyone,
As a guy who likes to self host everything from side project backends to multiple arr's for media hosting, it has always bugged me that for checking logs, starting containers etc. I had to open my laptop and ssh into the server. And while solutions like sshing from termux exist, it's really hard to do on a phone's screen.

Docker manager solves that. Docker Manager lets you manage your containers, images, networks, and volumes — right from your phone. Do whatever you could possibly want on your server from your phone all with beautiful Material UI.

You can get it on play store here: https://play.google.com/store/apps/details?id=com.pavit.docker

Key Features
\- Add multiple servers with password or key-based SSH auth
\- Seamlessly switch between multiple servers
\- Manage containers — start, stop, restart, inspect, and view logs
\- Get a shell inside containers or on the host itself (/bin/bash, redis-cli, etc.)
\- Build or pull images from any registry, and rename/delete them easily
\- Manage networks and volumes — inspect, rename, and remove
\- View real-time server stats (CPU, memory, load averages)
\- Light/Dark/System theme support
\- Works over your phone’s own network stack (VPNs like Tailscale supported)

https://redd.it/1omoxnn
@r_devops

check this cool vs-code extension I created

https://marketplace.visualstudio.com/items?itemName=AmitPatole.mcp-marketplace&ssr=false#overview

https://redd.it/1omtd40
@r_devops