How to transition in DevOps from Linux Administration?

I graduated and got a job as a linux system administrator have been doing that for the last 2 years. I am CCNA, RHCE and AWS Solutions Architect certified and am studying Kubernetes currently. How can I transition into a DevOps role. I understand, in terms of studying the technology, Theoretically, I am on the right track, but since my job profile defines me as a system admin, how can I crack interviews for DevOps job roles, realistically without taking too much of hit on my salary.

https://redd.it/1d1bg95
@r_devops

How do you manage versioning of your env?

I run cluster with 10 different microservices and frontend. How do you manage the versioning itself? How did you implement it? Should every service has its own version and should I use semantic approach or short hash or something else? We are using jenkins for our cicd

https://redd.it/1d1a1qp
@r_devops

Need some info on remote job

I am learning frontend web dev. If everything goes correctly in 5-6 months i am aim to get a remote job. Here are some questions i am really stressing about :
1- in remote job can take breaks btw hours according to my choice

2- as a beginner with no degree will i get a frontend web dev job…if i build good projects and have skills

3- my aim is to get minimum 66k usd a year….is it realistic expectation?

4- if i get a job in US or Europe will i have to pay taxes according to their rules ?

https://redd.it/1d1dnth
@r_devops

Does using chatGPT make me dumber as a devops engineer?

I'm using chatGPT daily to write automation/pipeline code and have achieved some success. It's save me the time from looking boto3/ansible API and writing that code. However I'm kinda worried that I will rely on it to much and can't interviewed well to land a next job. What do people think about this matter?

https://redd.it/1d1gpl0
@r_devops

Laid off, looking for a job


Hey everyone,

Last week, I got laid off from my position as a DevOps Engineer at Zeta (client project), and I'm currently looking for new opportunities. I have 2 years of experience in the field and am open to work and I prefer Bangalore location. If there are any openings in your companies, I would appreciate a referral. Thank you!

### My Experience and Skills:
- CI/CD: experience with Jenkins and ArgoCD for setting up and managing CI/CD pipelines.
- Deployment: Proficient in using Helm for deploying applications on Kubernetes.
- Observability and Monitoring: Utilized OpenSearch, Prometheus, and Grafana for observability and monitoring of applications and infrastructure.
- Cloud Infrastructure: Worked with Amazon EKS for Kubernetes cluster management and service mesh for microservices communication.
- Infrastructure as Code: Written Terraform scripts to automate the provisioning and management of cloud infrastructure.
- On-Call and Incident Management: Handled on-call duties, monitored systems, and resolved issues promptly to ensure high availability.
- Network Troubleshooting: Diagnosed and fixed network issues to maintain smooth operations.
- Scripting and Automation: Developed automation scripts in Bash to streamline repetitive tasks and improve efficiency.

### Project Highlights:
- CI/CD Pipeline Implementation: Designed and implemented CI/CD pipelines using Jenkins and ArgoCD, significantly reducing deployment times and increasing reliability.
- Helm Charts Development: Created Helm charts for Kubernetes deployments, enabling easy and consistent application rollouts.
- Monitoring with elk: Set up and managed OpenSearch clusters to monitor application performance, log management, and alerting.
- Terraform Automation: Provisioned and managed AWS infrastructure using Terraform. Implemented networking, compute, and storage resources. Technologies: Terraform, AWS (EC2, S3, VPC, IAM), Shell scripting, Prometheus, Grafana

Please let me know if there are any opportunities available in your companies.

Thank you!

---

https://redd.it/1d1gvzu
@r_devops

Job interview question

What are your steps to troubleshoot pipeline if the prod deployment from time to time failed and everything is working fine in other environment?

https://redd.it/1d1k4hn
@r_devops

Flat Files vs. APIs: How to Justify the Switch? (Need Arguments!)

Hi everyone, I work in a company where flat files are the norm. We're considering switching to an API-first strategy, but many colleagues don't see the point. They highlight the complexity of APIs and the lack of in-house skills.

However, we're encountering more and more problems with flat files: synchronization difficulties, handling errors, lack of data visibility...

I'd love to hear your feedback:

What limitations of flat files have you encountered?
What concrete benefits have APIs brought you?
How did you successfully demonstrate the added value of APIs to your team?
I'm open to any arguments that could help me defend this project!


https://redd.it/1d1m0x9
@r_devops

How to Access AWS From Azure VM Using OpenID Connect

Do you work in a multi-cloud environment?

Do you usually find yourself passing around cloud credentials? Hasn't it ever felt kinda wrong?

Did you ask yourself if there's a better way around this?

I'm here to tell you that there is. There is a much better way to handle such service-to-service communications.

This blog post elaborates on what OpenID Connect can do to help you avoid passing around long-lived credentials, relieving you from the chore of frequent secret rotation.

If you enjoy this post, please share it with your network.

#aws #azure #oidc #openidconnect

https://developer-friendly.blog/2024/05/27/how-to-access-aws-from-azure-vm-using-openid-connect/



https://redd.it/1d1mjhd
@r_devops

What to do with AWS consulting partner in devops team

Long story short, i’m about 4 to 5 FTE short in our team in order to deliver the projects the business wants.

I’m an acting lead and I’ve been asked how I might use two or three contractors from a AWS consulting agency to help us deliver on our BAU and devops uplift work.

My pain points are different levels of maturity across our cloud stack.

I have several EKS environments running just a single API application and what I feel is a very fragile terragrunt / helm stack (setup like 5-6 years ago) our api product hasn’t really evolved just waiting for onprem project to kick off next year but feel this infrastructure as code isn’t ready for it.

EKS upgrades are serious toil on my small team. By the time we get through the quarterly upgrade cycle we’re planning it again 8-12 weeks later. Unlikely to get buy in on switching to ECS due to onprem move planned next year.

I’ve got another 8 or so AWS accounts mostly serverless, ec2 pets in a mixture of cloudformation and SAM. I’d ideally like this moved to cdk to empower developer self service. Because my team need to help deploy new resources via cloudformation an a lot of clickops has snuck in.

My existing team are familiar with terraform and our terragrunt stack but would struggle with creating new modules or using a public one to implement something.

They know a bit of bash but not python or node so cdk will be a massive leaning curve imho (but I picked up enough typescript in 3 months to support cdk within my former team to support developers and using a bit of ChatGPT and cdk got some automation lambdas working!) as they more so come from onprem world and dabble on cloud/devops.

Assuming I have the opportunity to hire 2-3 devops engineers from an AWS consulting partner and they worked within my team how would I best use them?

I don’t have any infa patterns to hand over and I’m capacity constrained due to being a contributor to other projects as well as acting lead.

https://redd.it/1d1pjfe
@r_devops

It seems like I dislike every job

At the moment I have 3 years of experience in cloud and have felt always fairly motivated for my job untill my last 2 jobs.
Now I am already at my 4th company an I think I am switching jobs too often.

My current job I also get constantly annoyed by how weird organizations have decided to deploy software and infra. I have to deploy infra that makes uses of a 10k+ lines of terraform and I feel constantly lost and usually nobody awnsers any of my questions. Sometimes I think 70% of my team doesnt know what is going on in the templates.

Already I feel the urge to move to another job again, but this will look bad on my resumee. I have the feeling that it is about me, but I just really struggle usually with the way of working and feeling useless.
I am not a genius but I have the feeling I am not stupid either, I just get really frustrated really quick with dealing with other people's over engineered stuff.

Maybe I just need to be okay with not adding value yet after 1 month of working in this place. But I just have the feeling freelancing is more my thing. But for this my experience is still a bit lacking.

Now I am curious if other people also feel like they dislike all companies they have worked for. Do I just suck it up? Do people have advice of how to care less about the job?

Thanks in advance

https://redd.it/1d1qps7
@r_devops

Cut K8s costs with Karpenter

If you're like most companies running Kubernetes, you've likely seen your infrastructure costs balloon over the past year. Between supply chain disruptions, the rise in cloud prices, and increased app usage, those pesky cloud bills keep getting bigger and bigger.

It's enough to give any CTO heartburn. After all, you implemented Kubernetes and containers to increase efficiency and reduce costs! Yet somehow, it feels like you've lost control of your spend.

The problem often comes down to the limitations of your legacy cluster autoscaler. Tools like the Kubernetes cluster autoscaler were great when Kubernetes was new. But let's face it - their capabilities are limited compared to today's workloads. Relying on rigid auto-scaling groups and similarly sized nodes just doesn't cut it anymore.

The result? You end up with inefficient overprovisioning, idle capacity, and nodes that don't properly align with workload needs. And those inefficient nodes drive up your cloud spend exponentially.

Fortunately, there's a better way. Karpenter - the new open source node provisioning tool purpose built for Kubernetes.

Karpenter takes a workload-centric approach to right-sizing nodes on the fly based on actual resource requests. It also automatically consolidates workloads onto fewer nodes to minimize waste. Engineers who have made the switch cite 40% cost reductions or more!

Read how here https://www.perfectscale.io/blog/getting-the-most-out-of-karpenter-with-perfectscale

https://redd.it/1d1sc5l
@r_devops

From your source code to zero-downtime, high availability, and secure production deployment in no time (v5.0.1)

* With your project and its sole Dockerfile, Docker-Blue-Green-Runner manages the remainder of the Continuous Deployment (CD) process with [wait-for-it](https://github.com/vishnubob/wait-for-it), [consul-template](https://github.com/hashicorp/consul-template) and [Nginx](https://github.com/nginx/nginx).
* Examples in PHP, Java, and Node.js
* [https://github.com/Andrew-Kang-G/docker-blue-green-runner](https://github.com/Andrew-Kang-G/docker-blue-green-runner)

https://redd.it/1d1vbw4
@r_devops

Client wants to start storing data regionally

We have a client who sells services to other companies. We originally a NextJS application with PostgreSQL database in AWS Canada region.

They are looking to partner with a company in the US who seems to think they need user data stored in the US. But the current user data needs to also stay in Canada.

This presents many issues as the application was never designed to store multi-regional data like this.

One option is that we can deploy the tech stack in another AWS region, but then we essentially have two totally separate systems operating that don't talk to each other.

I've come across some service such as CockroachDB which has some features to store row level data in different regions but changing the database system at this point would be quite a bit of work. I'm not sure if something similar can be accomplished with RDS PostgreSQL.

I'm wondering if anyone has had to deal with a request like this before and hoping to get some ideas as to where to start.

https://redd.it/1d1xbbd
@r_devops

Are all CI/CD Tools similar?

So I started an intern position in a devops/IT role and am using Jenkins for the first time. Looking online, it looks like Jenkins gets a bad rep and it's alternatives are usually preferred.

So my question: are Jenkins and all the other tools similar in understanding concepts and usage? If I learn how to use Jenkins, would it be easy to use other tools?
I am mainly curious because I am hoping to learn as much as I can from this internship, and I hope that learning Jenkins means I understand how to use other tools to some degree.

https://redd.it/1d20ach
@r_devops

Preferred Dev/Deploy stack for basic python database app?

Retired devops guy here, almost a decade out of date. To update my skillset, I’d like to write a web/database app for a non-profit org I volunteer for.
My favored language is python. I’d like to do all-the-things as code, including provisioning, deployment, and database structure. I’d like to host it in the cloud, preferably free/cheap tier (it won’t get much traffic).
I’m starting the research phase, and thought y’all could help me focus in on what might be appropriate to look at.

https://redd.it/1d2cmj5
@r_devops

Ever Struggled with SSL/TLS Certificate Chains? Check This Out!

Hey everyone 👋,

I've just published a simple guide on SSL/TLS certificate chains on GitHub. This is my own notes I have kept in for a long time and I thought I shared to everyone and I'm curious to know what you think? I made it very simple enough to understand, breaking down what certificate chains are, how they work, and gave a real world example.

If you find it useful, please star or watch my repo. If not, any feedback to make it better and simple?

Check it out and let me know: GitHub - Understanding Certificate Chains: A Simple Guide

Thanks in advance!

https://redd.it/1d2dk01
@r_devops

Eraser AI

I found a useful tool today after I sifted through a post here from several months ago via my technical diagram query. I’ve worked in the Fortune 100 space for the last several years and diagramming tools have not been easy or fun. I don’t think the AI feature of Eraser was out several months ago when this product was discussed, but I think it deserves another discussion.

For those with other AI-powered technical visual/design tool experience, does Eraser AI detail a complex architecture better than the others, LucidCharts beta AI for example
? I tried diagramming with ChatGPT 3.5 in the beginning and wasn’t impressed. But models have clearly improved since then, and watching Eraser AI diagram my two year project in less than a minute with < 200 words was satisfying.

https://redd.it/1d2bzms
@r_devops

Begin for DevOps team

Hi all,

I'am a (m, 43y) I just finish some courses and have certificates. I'am from the Netherlands and want to start working in a DevOps team. But first i need to get alot of experience is my feeling. Could u give me some advice to start with. My experience is homecomputers and home networks, installing hardware, software and build computersystems.

Eager to start...

https://redd.it/1d2ewl1
@r_devops

I need 1000 concurrent iOS builds via Jenkins

We have a special requirement where we'll need 1000+ builds. All building concurrently.

Now, our main CI is Jenkins hosted in GCP.

What would be the best way to have this running?

1. AWS - Not very straightforward. But is it possible via ASG?

2. MacStadium - Seems the best option here since it's plugin has dynamic provisioning.

3. Anything else?


https://redd.it/1d2j9iy
@r_devops

What does everyone think of this recommended path I shared below?

https://raw.githubusercontent.com/milanm/DevOps-Roadmap/master/DevOps%20Roadmap.png

I found this roadmap today and wanted to hear some opinions about it. I have some fair knowledge about Git, JS development, Linux, and networking.

If I can dedicate at least 2 hours a day for this journey if not more, how long should I take on average to learn more about each topic? Of course there won't be an accurate figure. But from everyone's experience, on average, how long does it usually take? And how many projects should I involve myself with?

https://redd.it/1d2k6az
@r_devops

Landed first pure DevOps Engineer job

Just got the offer after 2 interview round. Super happy, can’t wait to start, been browsing this sub for a while, and now I can finally say

oneofyouoneofyouoneofyou

https://redd.it/1d2lvmh
@r_devops