Is there another word of term for "inner dev loop" that im just missing?

My question is: how do you guys refer internally to the "inner development loop" when you're working on things? This is a battle I've fought time and time again internally cause we call it staying in 'inner dev loop' but that's not reaaaallly a term that's widely known and it's not one that I really ever used before coming here.

Ok, so a bit of context, how we think about the inner dev loop is a cycle of activities you perform locally while working on a feature or bug fix (think- writing or modifying code, building the app, running/testing changes, debugging, commit yo code). Typically, the faster and smoother this loop, the more iterations you can make & its usualllly where most of us prefer to spend most of our time ya know?

On the flip side is the “outer dev loop” which encompasses the broader development life cycle (think -planning and task assignment, Code review, collabing with the rest of your team, CI/CD stuff, staging and prod release, monitoring). But I'm really more focused on how to refer to the inner loop.

More on how we refer to it here if you're curious for more context: https://thenewstack.io/hello-world-what-happened-to-the-inner-dev-loop/

Bu i really want to know what do YOU guys call this idea? I've scoured the internet trying to find a shorter, simpler term or way to refer to this idea and there really doesn't seem to be much out there. :P

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

Optimize your AWS / GCP infra with AI

Hey everyone, we built a tool that helps you optimize your cloud infrastructure costs using a combination of AI and static Terraform analysis. It’s only about a month old, so we’d love feedback from the community to see if we’re building in the right direction.

You can try it right away (no signup) at [infra.new](https://infra.new).

The agent has access to custom tools we built to help it catch many of the edge cases you’ll hit when prompting OpenAI / Claude directly.

Capabilities:

* Real-Time Cost Insights: See estimated costs as you update your Terraform configuration.
* Up-to-Date Docs: The agent automatically pulls in the latest Terraform docs before every code change
* Code Checks Built In: A language server flags errors and feeds them back into the model for fixing
* Expert Human Examples: Our agent pulls code snippets from expert human-written examples
* GitHub Integration: Import your existing Terraform, see what it currently costs, ask the agent to look for optimizations, then export any changes back to GitHub

I’d love to hear your thoughts!

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

White label PM system

I offer white label SEO services and I need a better system to manage my “sellers” who have clients under them. Any recommendations? Thanks!

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

Sonarqube Developer to on premise Azure DevOps Server unable to find valid certification path to requested target

I started a Sonarqube Developer instance using the docker image:

sonarqube:9.9.8-developer


On the other hand I have an on premise Azure DevOps Server 2020.0.1 (Dev18.M170.8).

This Azure DevOps server runs on https, and it asks for Windows Integrated Authentication.


Sonarqube configuration:

In Sonarqube, I go to Administration > Configuration > DevOps Platform Integrations, and I add an Azure DevOps configuration:

Azure DevOps URL: https://ourServer..com/ourCollection/



Certificate configuration:

In the browser, I went to https://ourServer..com/ourCollection/ and I have exported the certificate to: myAzureDevOpsServer.crt


Then I've imported that certificate into the docker container Java Truststore, with this command:

keytool -import -trustcacerts -alias devops-cert -file myAzureDevOpsServer.crt -keystore $JAVA_HOME/lib/security/cacerts -storepass changeit


I've checked that the certificate was imported.



Behaviour expected:

Sonarqube retrieves the information about the projects in the Azure DevOps Collection.


Actual behaviour:

In the docker container logs I get:

2025.01.31 22:10:25 ERROR webAZS+X3ieDsjXs89xAABBo.s.a.c.a.AzureDevOpsHttpClient Unable to contact Azure DevOps server for request https:/ /ourServer..com/ourCollection/_apis/projects?api-version=3.0: PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderE xception: unable to find valid certification path to requested target



I'd appreciate any help. Thanks.

https://redd.it/1ier4jt
@r_devops

Managing Azure Policy with IaC at scale (Discussion)

Hi guys, curious to know what methods out there in the wild you're using to manage Azure Policy deployment via Terraform?

Especially for enterprise environments where you have multiple policy assignment scopes, management groups, subscriptions, resource groups.

Whilst using built in and custom policies and policy initiatives. I'm currently architecting a solution for Azure policy governance at scale with Terraform, and keen on knowing the various approaches out there !

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

How much software development should I be able to do?

I've never been able to figure this out. What level of software development experience should I have for a typical Platform engineer job? Usually when I write code, it's just Python or PowerShell, and declarative languages. I don't think I could write a fully working piece of software myself, but at this point I'm not sure how to get to that point without a CS degree. Everything else I learned from KodeKloud and lots of lab time. My background is as a systems engineer though, so obviously not as smart or talented as a SWE.

https://redd.it/1iet8jt
@r_devops

Do you think we are doing Blue-Green deployment?

I've been reading and watching blue-green deployment solutions on different webistes as well as youtube. The more I watch and read different articles, the more it's telling me that our implementation does not do blue-green deployment.

This is the process or flow of ours.

1. Spinnaker creates ASG(let's name it A), load balancer and EC2 instances are created based on the ASG launch template

2. If developers wants to make a new deployment, they update their git project and submits a merge request

3. Spinnaker creates a brand new ASG(let's name it B). It DOES NOT create a brand new load balancer. EC2 instances are launched from the new ASG which is B

4. If the applications on the freshly deployed EC2 instances(created by ASG B) is healty, all freshly deployed EC2 instances are registered to the LOAD BALANCER

5. Previous EC2 instances deployed by "ASG A" are deregistered from LOAD BALANCER and are TERMINATED.

Based from the articles I've found and read, BLUE-GREEN deployment does not deploy everything right away and does not terminate resources right away. It's like percentage based. However, our BLUE-GREEN deployment terminates all the previous resources right away when fresh deployment is healthy.

Otherwise, when freshly deployed EC2 instances are not healthy, it doesn't get registered to the LOAD BALANCER nor the current running EC2 instances are terminated.

Is ours considered a BLUE-GREEN deployment?

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

How can we factor non-standard things in terraform deployments

Like if some NW parameters is set allowed for all connections but terraform will allow deployment. So how can we avoid it trigger this beforehand.

https://redd.it/1iex3c6
@r_devops

Is devops good for my future?

I just got assigned to devops in my internship, I am in my final year. Basically a backend type guy.

Questions -
1. Do devops guys write real code?
2. How important is devops from the view of a senior leader?
3. Does it just involve writing yaml files?
4. Will my critical thinking ability be polished here?
5. After mid life will I still find a job?

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

CDTC approach

I just realized I have been using it for the last 10+ years and never invented a name for it. When we have infra code, we want to test it (for real) before deploying it in production. Because it's infra code, we need ephemeral infra subjects (usually virtual machines, bare-metal servers, or switches) to deploy it to before testing. We can't get away with unit tests or some 'in the docker' silliness.

And every good infra test has a pattern of 'create, deploy, test, destroy'; hence, CDTD. (I made a typo in the title and I can't change it).

Is it so? Am I too late to the naming party? How do you call this approach?

https://redd.it/1if1yev
@r_devops

jenkins first steps

Hello, I'm new to DevOps. I have knowledge of Docker, Kubernetes, and scripting, and I want to take my first steps with Jenkins.

I have read the Jenkins documentation and watched a few brief YouTube videos about the GUI and initial configuration.

I need to complete a final project to get my degree. Although I have already worked on other projects, I want to start my first DevOps-related project because my goal is to work in this field.

Do you have any course recommendations or video tutorials to learn Jenkins? I'm willing to read all your suggestions!

Also, feel free to share any recommendations about career paths, certifications, or technologies that are not the main focus right now but are worth learning.

Thanks in advance!

https://redd.it/1if4qle
@r_devops

How to upload a Lambda function with Node.js SDKs and dependencies?

Hello, I have a Lambda function (index.mjs) file that relies on a few SDKs and dependencies to run. The function performs the following tasks:

1. Retrieves files from an S3 bucket.
2. Uploads them to an APS OSS Bucket.
3. Returns an URN.

I’m trying to figure out the best way to upload the index.mjs file along with its Node.js modules and dependencies (like AWS SDK, etc.) to the Lambda function.

What’s the proper approach for packaging and uploading this Lambda function with its dependencies?
i have tried zipping all the contents from local and have uploaded it inside the lambda function
but i'm constantly getting some node module errors.

suggest some advice or best practices which would be very helpful for me to achieve this process.

Thanks!

https://redd.it/1if4rzc
@r_devops

Moving Away from Jenkins ... ...

In my group - i've setup Jenkins from scratch (bout 8years ago give or take) , , went through the pains and good times of my time with Jarvis - lol .. .. But now im on old servers and getting the nagging emails from the ultimo PMO group "your servers are out of scope - unsupported - out of compliance"

I get it - yes im very late and all - - but wow PMOs , , damn - lol .. what do they actually do - - w/e thats another topic I guess.

so - im moving away from JENKINS I feel - - its too much to scale , DR and all - - SO, github actions GHA anyone - lol ..

but im getting mixed feelings and reviews on transistioning my jenkins jobs - - which are app deployments - - and a mix bag of deployments

= = = some are just files delivered from the vendor where I use Jenkins to do backups , file copies and such

= = = some are just "jobs" I have that run sqls to PUSH or PULL data to an ftp , s3 bucket

= = = and some do a full ci/cd where it checks out from github , does an ms build and copies to servers

pretty much all done via various python / powershell scripts. and im using JENKINS as the orchestrator , scheduler and really a single point for all my "jobs"

now will GHA help me out - - is it worth the transistion?

https://redd.it/1if9aiy
@r_devops

What’s your approach to assessing velocity using ADO (Azure DevOps)?

We’re using Fibonacci and have epics>features>PBIs>tasks.

We’d like to assess velocity to launch, but our major go live is next January. We don’t have all full backlog.

Would love input. Thanks.

https://redd.it/1ifbg45
@r_devops

AI-Powered Secret Scanning

Hi everyone,

I'm working on a new secrets management platform for developers called Stashbase, and I’m working on one feature: AI-powered scanning for hardcoded secrets in your code. This tool can run checks during your pre-commit or pre-push Git hooks, or as part of your CI/CD pipeline.

The tool intelligently detects hardcoded secrets in your code and prompts you to either remove or sync them with a secrets manager (like Stashbase). You’ll also have the option to configure which files to ignore, set up custom scanning rules, and more.

I’d love to hear your thoughts:

Would you use a tool like this?
How important is it that the tool is intelligent and customizable?
What are your thoughts on running scans before committing or pushing your code?
Any features you’d love to add?

Looking forward to your feedback! Thanks for your time!

https://redd.it/1ifew0e
@r_devops

How Do I Get Back Into IT After a 2-Year Gap?

I graduated in 2023 with a B.E. in CSE. After that, I worked on client projects and built personal apps using React Native and Next.js. I also deployed Node.js and Python apps using Coolify, Docker on DigitalOcean Linux machines, and Railway.app.

For the last 2-3 years, I had serious health issues and couldn’t actively look for a job. Now, at 25, I’m finally healthy and trying to get into DevOps, but I feel like I’ve lost valuable time. I’ve applied for several jobs, but I keep getting rejected.

Since I’m based in Bangalore, I heard Azure is widely used here. Should I focus on Azure over AWS, or should I get certifications for both? Also, is Terraform worth learning?

What should I prioritize first? Are there any beginner-friendly DevOps projects I can build to stand out? And how do I handle the employment gap when applying for jobs?

Would really appreciate any advice or a roadmap! Thanks in advance.

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

₹ 1Cr in DevOps: Myth or Achievable Reality?

Hi Community!

I’m curious—how realistic is it for a DevOps Engineer in India to earn a ₹1Cr salary? What kind of exceptional skills would one need, and what’s the typical years of experience required to reach this level? Would love to hear insights from those who have seen or achieved this milestone!

https://redd.it/1ifs8w4
@r_devops

Gus need help! How do I pull such huge logs from loki ?

As the title says, I have a service that logs 8 Million lines / Min. I should be able to pull minimum 5 minutes of logs.
Single API call can give me only 10k log lines. Linear pagination would take forever. Multithreading or asyncio connections are throttling queriers. I can scale up queriers though (100 currently) but there has to be an efficient way

I personally thinking to implement kafka or rabbitmq but not sure if that would actually solve problem or would just be an overhead

Any experienced peeps with a better solution ?

https://redd.it/1ifuec1
@r_devops

Full Stack Developer Looking to Transition into DevOps – Need Guidance!

Hey everyone,

I’m a MERN stack developer with 2 years of experience, but I’ve been finding it incredibly difficult to get clients due to the highly saturated market. Because of this, I’m considering a move into DevOps, but I’m unsure where to start.

I already have a solid understanding of computer networking concepts and related topics, but when it comes to AWS, I feel overwhelmed by the number of services and certification options available.

🔹 What should be my first step in AWS?
🔹 Which AWS certification is best for beginners in DevOps?
🔹 What skills should I focus on to become a strong DevOps engineer?

I’d really appreciate advice from those who’ve transitioned into DevOps or are already working in the field. Any guidance, resources, or learning paths you recommend would be super helpful!

Thanks in advance! 🙌

https://redd.it/1ifzc8r
@r_devops

We made an open source testing agent for UI, API, Visual, Accessibility and Security testing

End-to-end software test automation has traditionally struggled to keep up with development cycles. Every time the engineering team updates the UI or platforms like Salesforce or SAP release new updates, maintaining test automation frameworks becomes a bottleneck, slowing down delivery. On top of that, most test automation tools are expensive and difficult to maintain.

That’s why we built an open-source AI-powered testing agent—to make end-to-end test automation faster, smarter, and accessible for teams of all sizes.

High level flow:

Write natural language tests -> Agent runs the test -> Results, screenshots, network logs, and other traces output to the user.

Installation:

pip install testzeus-hercules

Sample test case for visual testing:

Feature: This feature displays the image validation capabilities of the agent Scenario Outline: Check if the Github button is present in the hero section Given a user is on the URL as https://testzeus.com And the user waits for 3 seconds for the page to load When the user visually looks for a black colored Github button Then the visual validation should be successful

Architecture:

We use AG2 as the base plate for running a multi agentic structure. Tools like Playwright or AXE are used in a REACT pattern for browser automation or accessibility analysis respectively.

Capabilities:

The agent can take natural language english tests for UI, API, Accessibility, Security, Mobile and Visual testing. And run them autonomously, so that user does not have to write any code or maintain frameworks.

Comparison:

Hercules is a simple open source agent for end to end testing, for people who want to achieve insprint automation.

1. There are multiple testing tools (Tricentis, Functionize, Katalon etc) but not so many agents
2. There are a few testing agents (KaneAI) but its not open source.
3. There are agents, but not built specifically for test automation.

On that last note, we have hardened meta prompts to focus on accuracy of the results.

If you like it, give us a star here: https://github.com/test-zeus-ai/testzeus-hercules/

https://redd.it/1ig0lfk
@r_devops

I'm a software engineer, should I take a DevOps job?

I'm a software engineer at a consultancy that requires DevOps as well. I'm thinking about taking a job that is out and out DevOps. I enjoy DevOps/platform work but what makes me slightly unsure is do I want to do it full time and give up writing software. Are there and software engineers that made the switch to DevOps? If so, do you have any regrets or is it all positive?

https://redd.it/1ig1ccx
@r_devops