Kubernetes is the new Jenkins....

With all the operators, extensions and nightmare of keeping it up to date kubernetes is the new Jenkins....

https://redd.it/1imk1ry
@r_devops

NOC to DevOps or Cybersecurity?

Hey guys, a little bit about myself, I'm turning 22 soon, I've started my career through Tier 2 Support Specialist for 2 Years, made a transition to NOC within the same company (Present) currently I'm in this position for 4months.

This department will be closed In May and I don't know which path to choose.

NOC > Junior DevOps or other similar rules, (Existing experience In DevOps tools and cloud while I keep learning during my work)
NOC > SOC / Incident Respone Analyst (where I need to learn from scratch but I've always had passion for ti)

If you guys were in my shoes, which path you'd choose or what you'd do if u were me?

If there are any more similar rules feel free to list them here.

Thank you guys it means a lot!

https://redd.it/1imkv5l
@r_devops

Anyone cares for Datadog Vector as SaaS?

I see developers and DevOps struggling with running in-house OTel pipelines.
Would you like to subscribe to a SaaS version of Vector (https://vector.dev/)?

The only prohibitive cost would be the outbound cost, which can be offset if the SaaS service provides a CloudFront link to send the data to since the cost of EC2 to CloudFront is USD Zero.

Would you still not use this service, and why?

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

How to Publish to GitHub Pages From Another Repository

Hey DevOps folks!

I wrote a detailed guide on deploying static sites from one GitHub repository to another using GitHub Actions and OpenTofu.

This setup is particularly useful if you want to:

- Keep your source code private while using free GitHub Pages hosting
- Manage infrastructure as code using OpenTofu/Terraform
- Automate cross-repository deployments with GitHub Actions

The guide walks through:

1. Setting up the target GitHub Pages repository
2. Configuring the source code repository
3. Creating necessary deploy keys and GitHub Actions workflows
4. Implementing the deployment pipeline using OpenTofu
5. Managing the infrastructure with Terragrunt

All code examples are provided, including complete GitHub Actions workflows and OpenTofu configurations.

https://developer-friendly.blog/blog/2025/02/10/how-to-publish-to-github-pages-from-another-repository/

Let me know if you have any questions!

Please share in the comments if you prefer an alternative approach.

https://redd.it/1imb0fy
@r_devops

YAHH - Per-project history file

YAHH is a Zsh-based tool that helps you manage separate command histories on a per-project basis. Instead of having one global history file or one per directory, YAHH allows you to keep distinct histories—called realms—for each of your projects.

This makes it easier to recall recurrent commands that are specific to a given project or operational environment, useful in professional services, consulting and other context-switching role.

https://github.com/Positronico/yahh

https://redd.it/1imt4fz
@r_devops

WebUI for Ansible with some monitoring?

A while ago, I used Puppet and Foreman, and it was quite practical. I used it for the whole pipeline, from setting up hosts to configuration management. However, I now use Ansible more and more, and I find that Foreman doesn’t integrate well with Ansible. Also, Ansible Tower/AWX is not very comfortable to use. So, I mainly rely on the command line, which works well enough. That said, I was wondering if there is a modern GUI for Ansible that also includes basic monitoring features, similar to Foreman, where Facts where visualized.

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

Implementing deployment windows

Hi,

I've been asked to look into managing some code releases (across various systems and methods) through specific pre-arranged release windows and I'm trying to find information about good ways to do this, pros and cons etc. However searching online is a problem because I just get endless results about deploying MS Windows changes 😂

Can anyone point me in the direction of some good resources or feel free to reply with opinions? I'm obviously aware this is going to work better for some types of change than others.

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

List of 650+ well-funded startups that don't suck (remote, US, EU)

Finding well-funded, growing startups with strong engineering/product cultures is really hard. Created www.startups.gallery to make finding them easier. And no, this is not another spreadsheet or pay-to-play directory. It's just a thoughtful collection of today's most interesting projects, curated by humans. And yes, I know that startups aren't for everyone, but these are hopefully the most promising ones. Open to all and any feedback!



https://redd.it/1imzow5
@r_devops

Confused and scared please help!!

Hi guys
I am an AWS Architect with pretty good experience. I recently switched companies. I had interviewed for an Architect role. I got assigned to a project and it's AWS Devops. I got the initial KT and feel like a fish out of water. I am scared if I can cope up and do well. Please help me I am literally going thru mental breakdown....
I don't know if I should ask for another project or what?!

https://redd.it/1in1w4n
@r_devops

How would you architect an an analytics webapp?

I architecting an app that will have terabytes of ecommerce product data (products, titles, review) that is being scraped every 15 minutes and stored into S3 as JSON. The clients in this case will be a web app (mobile app in the future) that shows trends over time (such as price trends), number of reviews, etc. Users will be able to select date ranges and other ways to filter and aggregate the data. Frontend is written in React, backend and pipelines are written in Python.

I am torn between the following:

1. just simply doing ETL from S3 into TimescaleDB, which the apps will query.
2. doing ETL from S3 into Clickhouse for analytics, and a separate Postgres instance for relational queries (users accounts)
3. Implementing a lakehouse architecture by using Apache Iceberg on top of S3 and seeing if the app can use a Python Trino client to query. Lakehouse architectures seem more efficient than traditional data warehouse architectures - I'm just unsure of the performance within applications.

How would you build this?

https://redd.it/1in1n93
@r_devops

I ask ChatGPT about AI replacing DevOps engineers every quarter and this time it finally gave me an answer

Question:

>How long does it take till AI replace the majority of devops and sre engineers. Give me a number not the stuff like AI won't replace bla bla bla

Answer:

A realistic estimate is **5 to 10 years** for AI to replace the **majority (over 50%)** of DevOps and SRE engineers.

* **1-3 years**: AI will significantly **augment** DevOps/SRE roles but not fully replace them. Expect automation of routine tasks (CI/CD, monitoring, incident response).
* **3-5 years**: AI will handle **most** of the operational workload, reducing team sizes. Companies will rely on AI-driven platforms for reliability engineering.
* **5-10 years**: AI will likely **replace** more than **50% of DevOps/SRE engineers**, keeping only a small number of experts for oversight and edge cases.



https://redd.it/1in6ual
@r_devops

How Do You Track Developer Environment Costs?

Hiiii, so I'm doing some research on how companies approach developer budget costs, specifically around environments. Sorry, feel free to brain dump, just working to understand how others might do it differently than me.

For those of you managing a dev team and balancing environment costs—do you track expenses per developer or per environment? Do you assess costs on a monthly or yearly basis? And do you try to maintain a certain ratio between staging and production costs?

For example, at my company (on the small side) we track staging vs. production costs monthly but don’t break it down per developer. I’m curious if larger companies handle this differently—or if other small companies do things another way.

Hit me with your thoughts :)



https://redd.it/1in7svm
@r_devops

Does anyone have a AWS CDK example of a ALB + ECS Fargate + CodePipeline blue/green setup?

I am really struggling to find a holistic example of this in documentation or elsewhere. I'm CONSTANTLY running into a chicken or the egg scenario between ECS and CodePipeline. In click-ops I can get it working almost instantly but its proving to be a serious pain for me in my AWS CDK IaC project. Feel like I've tried a million combos but nothing has worked E2E yet.

Note: I'm talking about a full ECS Fargate + CodePipeline (+ source, build, deploy) setup btw - where we have the task defs/appspec in the source repository, then want to fetch and use them as well as ECR image during each pipeline execution.


I know this is from a knowledge gap on my end, any help is super appreciated.

https://redd.it/1in90we
@r_devops

Redshift Disk Space Maxed Out, Crashed Processes – Unexpected Storage Lock?

Our Amazon Redshift cluster’s disk utilization suddenly hit 100%, taking down key processes. No obvious culprits—no excessive logs, no runaway queries, nothing abnormal in monitoring.

After some troubleshooting, we ran a VACUUM, which immediately freed up space. It seems like something locked unused storage and wasn’t releasing it—possibly a storage system glitch or an unexpected background process.

Still unclear on the root cause.

Screenshot here.

Has anyone else run into this with Redshift? Any insights? Thank u!


ps. not seeing anything on X(twitter) , and the aws status page does not indicate any issues .

https://redd.it/1in8m6w
@r_devops

Can anybody let me know where is the entire CI/CD Pipeline Implementation tutorial in KodeKloud? I am unable to find specifically for CI CD Pipeline.

.

https://redd.it/1indd4c
@r_devops

Is there a way to prevent you from committing into master branch in your local?

Is there a way to prevent you from committing into master branch in your local? I use husky, and I am wondering if there's a way to prevent me from committing into master branch locally, because sometimes I forget I am on master and I commit before creating a new feature branch.

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

Is it even practical to switch from IT operations to SWE (devops or Cloud) role at 36

As the title says I am 36 and in IT operations where I am working in same project for last 10 years. I feel I am at dead end and have nowhere to go. Not doing any technical stuff other than restarting servers, some Linux commands ,updating bash scripts, setup splunk and new relic monitoring. Late last year I have decided to switch to devops or cloud role and completed Google IT automation with python specialization on course era and also took aws solutions architect certification.Started doing leetcode this January but I am not sure what kind of positions I can apply as I am not fresher or don't have experience they expect for a devops or cloud engineer. I can't start fresh as I have family responsibilities and need to earn at least what I am earning now(140k+USD) .I am kind of lost on how to proceed ... And rethinking if it's even practical for me to switch at this age? Posting here for any guidance on how to prepare for the switch specially what kind of projects I can do to show some experience and explain in interviews ?(Any course or websites or git links)I am ready to spend whole year preparing at least 10 hrs a week .

Edit: Internal move is not an option as I am a subcontractor at a federal project

https://redd.it/1inisy5
@r_devops

Do you have a dedicated engineer (s) who's job is looking at metrics, monitors and alerts?

I understand this should be part of each engineering teams job or the SRE teams job. But the engineers are too busy to improve or maintain monitoring or alerts.

Is there a role in companies who's sole job in looking at dashboards, adding monitors and alert workflows, etc.?

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

Challenges with Real-time Inference at Scale

Hello! We’re implementing an AI chatbot that supports real-time customer interactions, but the inference time of our LLM becomes a bottleneck under heavy user traffic. Even with GPU-backed infrastructure, the scaling costs are climbing quickly. Has anyone optimized LLMs for high-throughput applications or found any company provides platforms/services that handle this efficiently? Would love to hear about approaches to reduce latency without sacrificing quality.

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

Is it possible to change to dev role after devops?

Hi members. I am a 2022 grad. I got into devops as I was not getting calls for dev role and I had done an internship with a fintech as SRE. So I applied for a devops associate role. Fortunately cleared all rounds and got selected.

Even to this day, I get assigned work that is related to depooyments and deployment scripts etc and some tasks related to kubernetes. I also work on some docker related parts where we have to create docker images for some jenkins nodes that we run.

We use AWS and GCP here. But I get no task assigned for those. I have talked with my manager he says to have certification or atleast complete courses so that I can assign those tasks to you. I am trying to follow courses but not very consistent.

As to why I am not that consistent is because honestly I think I don't enjoy devops anymore or atleast the devops tasks that I am getting assigned here. Most of the distaste comes when I see that developers having the same experience as me get paid higher as high as 66% more than me. Before joining devops I was a CS grad and I had done a lot of coding and built applications etc. So I have knowledge atleast.

So I want to change to a dev role. Its not easy to get hired as a fresher in my country because most of the fresher hiring goes through on campus placements. Unfortunately I also got placed in a company as dev from my college but that company didn't had much work for me. I was not assigned to any projects for 8 months. I still hate that company. And that is the reason why I wanted to change. Change to a different company as a developer was not possible as I had no experience. And started applying randomly to SRE and devops roles also. And that is how I landed in my current company as devops engineer.

So current situation is that I have 2 yrs of experience but its devops and I want to change to a dev role. I have done some personal projects relevant for dev roles like web applications,REST apis, android apps.

Can I shift to a dev role given my current situation or continue as devops? The recruiters look for "relevant" experience. Can I fake experience in my resume. Like can I tell them that I have 2 yrs of experience in my currently company as a dev? Can I list the tasks and responsibilities of devs in my current company in my resume to fake. Will they find out if I clear interviews?

https://redd.it/1inm6zj
@r_devops

Do you use Grafana On Call?

My team (4 people) wants Grafana On Call setup instead of the traditional Alert, Contact point, notification policies.
We use Grafana Open Source.

Their reasoning is a bit loose with why we should do this. It seems like it's a case of Grafana On Call just being a more modern, versatile tool. With their argument being that it can send us SMS messages.
Although, I've made the case that we can do more advanced alerting in Grafana or using Prometheus Alert Manager, they're insistent that we use Grafana On Call.
We do not work in shift patterns. Just regular 9 to 5.

I've been trying to set it up, but have just found it to be a bit of a faff. I've had to setup a Grafana Cloud account, as that seems necessary to allow certain features. Which means we've now got another location which our Grafana server pulls in information such as user details.
I've had to spin up several containers for Grafana On Call and it's redis server.
There's networking which has needed setting up around that.

Certain things I've had to troubleshoot and just found no mention of it within the documents, but random hints on Stack Overflow, etc.
I'm just wondering what everyone's opinion is on the service? Am I just unnecessarily whining and make things harder for myself?

https://redd.it/1inmnp0
@r_devops