Reddit DevOps
289 subscribers
94 photos
32.8K links
Reddit DevOps. #devops
Thanks @reddit2telegram and @r_channels
Download Telegram
PSA: Implicit TLS on port 465 is the future, not STARTTLS on 587

Basically, someone misinterpreted an RFC (particularly section 3.3), and now.

I for one was thinking that 465 was the "older" solution (mistaking it for SMTPS), so was reassuring to find this. Way easier to proxy TLS (thanks, SNI and ALPN!), so I'm happy to find out I was wrong.

Another post for the blogoverse.

https://redd.it/zc1mar
@r_devops
Want a playground

I want to build a Cicd pipeline and deploy apps and automate that. When I did this in aws my billing was so high and I couldn't afford. Is there any free website or something to practice building cicd pipeline? TIA

https://redd.it/zc4pzt
@r_devops
Question about cognito and oauth2 resource server

Hello,

In my company, we use api gateway with aws cognito.Our process is to create an API, add the swagger on API gateway and create a resource server with desired scope.We have our appclient accessing the needed scope and that's all.For one API we create, we create one resource server with its scope.

AWS limits the number of resource server to 25 (with a maximum of 300) within a single user pool. So in my case, i can only create 25 api until i get blocked.

Oauth2 definition of resource server is "The resource server is the OAuth 2.0 term for your API server" and the auth0 is "the server hosting the protected resources. This is the API you want to access". But it's not quite the case with cognito as it's "a remote server that authorizes access based on OAuth 2.0 scopes in an access token."

My question is : is my approach the right approach concerning the resource server within cognito ? or should i factorize my resource server somehow ?

https://redd.it/zc750x
@r_devops
Week 1 Jr DevOps Engineer

Introduction:
- I plan to discuss weekly what I’ve been getting up to as a junior devops engineer and the kinds of things I’m actually doing. This should be helpful and would love to get feedback if other devs use different technology or have suggestions for different things that I mention. I find this is a good way to hold myself accountable for what I’m producing.
- At the end of my juniorship then I will turn my weekly posts into a guide so others know what to expect.
- Please note this is a startup and so a lot of practices won’t have been implemented yet, this means I get to learn processes start to finish.
- The stuff I work on is highly confidential and therefore will be less specific.

Background:
- apprentice DevOps engineer 1 YOE large financial technology company.
- at the fintech I worked in doing mostly automation via Jenkins for AWS and managing cloud environments but also included working with reactJS to create some front end tools
- interview process was quite good with one design infrastructure question.

==Week 1==
- was told not to expect to do much other than onboarding for the first week but my brain is on steroids so I did all the onboarding in two days.
- familiarised myself with basics of kubernetes and accessed clusters and especially monitoring type stuff and azure
- made some yaml Prometheus type monitoring rules and set them up based on different metrics
- made my first pull request and merged after testing and approval
- created graphs and a nice dashboard to display these metrics using grafana
- note that during this time I was trying to still get all my stuff setup on work servers and get my laptop (now multiple because they didn’t add linux to my boot :O)

How do I feel:
- I feel pretty happy with how my week went, everyone is lovely and I feel like I’m in a really great place
- there are currently some bad practices going on and I have raised this, I will plan to create a PowerPoint to have these changes implemented and present it to the higher ups. My manager knows and really wants things to change too to make everything scalable.

https://redd.it/zc8zys
@r_devops
So - how did you get your first devops job?

Yo, bro, it's all about networking and experience, man. I started out learning the basics of Linux and system administration, and then I got some internships and entry-level jobs to gain some real-world experience. From there, I started attending meetups and conferences, and I networked like crazy.

Eventually, I landed a job as a junior devops engineer at a startup. It wasn't the most glamorous job, but it gave me a ton of hands-on experience with all kinds of cool tech. I learned how to set up and maintain CI/CD pipelines, how to deploy and manage microservices, and how to monitor and scale production systems.

I also spent a lot of time learning new technologies and keeping up with the latest trends in the devops world. I read books and articles, watched videos and tutorials, and experimented with new tools and frameworks in my spare time. This helped me stay up-to-date and relevant, and it also made me a better engineer.

After a few years at the startup, I started applying for other devops jobs and eventually landed a position at a bigger company. This job paid better and had more opportunities for growth and advancement. I continued to learn and grow, and eventually I was promoted to a senior devops engineer.

Now, I'm in charge of a team of devops engineers and I work on some of the most challenging and exciting projects at the company. It's been a long journey, but it's been worth it.

So, the bottom line is, if you want to get a job in devops, you need to start learning the basics and gaining experience. Then, you need to network and stay up-to-date with the latest technologies and trends. And, most importantly, you need to be passionate and dedicated to your work. Good luck, bro!

https://redd.it/zc9o5x
@r_devops
"If you control access, you control compromise..A very, very simple premise of Zero Trust is that everything will be compromised at some point. In reality, it’s everything is compromised until proven otherwise"

> "If you control access, you control compromise," said Dr. Chase Cunningham, a leading Zero Trust security expert, at a recent event hosted by secrets management platform Akeyless. "A very, very simple premise of Zero Trust is that everything will be compromised at some point. In reality, it’s everything is compromised until proven otherwise."

Source

I think this is a logical and pragmatic approach to security. In today's highly interconnected and complex computing environments, it is virtually impossible to prevent all security breaches. Therefore, it is important to focus on controlling access to sensitive assets, so that even if a breach does occur, the damage can be minimized. By adopting a zero trust mindset, organizations can take a proactive approach to security and be better prepared to respond to potential threats.

What are your thoughts on this take?

https://redd.it/zcalgl
@r_devops
Gathering experience in a stable infrastructure

I am a fresh devops engineer working for 5 months. It is a software company with about 500 employees including 120 developers. I am confused what to do here. The deployment, pipeline for CI/CD, aws instance, kubernetes cluster all are well mature and no one touch them. My teammates just do a little maintenance task. In this scenario, how can I grow my career or gather experience?

https://redd.it/z7z6cw
@r_devops
At what point does it make sense to deploy K8s?

As I'm learning about serverless, it seems like K8s is a big buzzword these days. Every job post out there wants K8s experience.

If I were capable of building a scalable web app using Cloudfront, S3, API GW, Lambda, and DynamoDB, at what point does it make sense to move from that dichotomy to something like K8s? What are some of the benefits of K8s that are not available when using the aforementioned services? More complex compute requirements that can't be handled by Lambdas? Or maybe your app requires super fast processing time that a container can do, whereas you'd need to wait for an async call to a Lambda?

I just don't know enough about K8s to really understand the use case, and I feel like so much of today's DevOps landscape is "just do K8s!" without really asking why.

https://redd.it/zcfmv3
@r_devops
What is helpful to you when someone is trying to find out your perspective on workflows?

Hi all,

I'm working with my developer team to get a current picture of the departments workflow and methodology as it's being followed and used. Including their pain points, what they like, what their dream processes might look like.

I'm wondering if you've ever been in this situation and if so: what questions helped to spark more discussion or conversation? What did you appreciate that happened...or what should someone like me avoid in these convos?


My goal is to get a sense of the as-used workflows and to eventually (months down the road) be able to implement changes/increase cross-department teamwork. I've been on the dev side before and generally when people come in to do this it doesn't feel very welcome or like it'll result in something helpful. I'm trying to avoid that from the get go and make it clear that I'm really here to work for the team.

https://redd.it/z87zvh
@r_devops
Bare metal -> VM -> Kubernetes setup

We have an outdated stack for kubernetes cluster. What would be a better/more modern solution to setup physical servers, VMs and a Kubernetes cluster?

Actual state and issues:

\- brand server (XCP-NG OS installed manually)

\- VMs on these servers (maintained by Xen Orchestrator)

\- kubernetes nodes on dedicated VM-s (via terraform + ansible + k3s)

Unfortunately Xen Orchestra does not have Packer support, so templates can't be upgraded easyly. And XO's Terraform is a bit complicated solution.

So we would like to replace/modernize the stack, there are a few things which cannot be changed:

\- we need VM-s, we have dedicated databases VMs, and we have a few legacy app which cannot be migrated to k8s

\- we need a working terraform/packer.io support

​

Do you have any suggestion what stack would be ideal for us?

​

Thanks

https://redd.it/zchqus
@r_devops
How to properly plan and align on big weekend maintenances?

The company where I work has monthly weekend maintenances. We create these Excel sheets where all of the involved teams (number is between 5 and 15 teams involved, with multiple interdependencies) write their actions: basically What, When and overall comments. Example: "Stop service XYZ at Dec/01, at 10pm".

The creation of this spreadsheet is manual and error-prone. Usually someone will copy the sheet from a previous maintenance, and change/adapt information. But things slip by, or old info is left there.

How can we do better?

Before anyone asks: for many, many reasons outside the control of my department, it is not possible to change the way we work. For now we'll still have monthly weekend maintenances. Within this context we have to do our best.

https://redd.it/z7w55j
@r_devops
CKA exam course

Hi,

What are some good materials (even paid) that I can train for CKA exam?

Thanks

https://redd.it/zcl5dt
@r_devops
Learning materials for CI/CD stuff as a data engineer

What are the good learning materials that you can recommend in learning CI/CD stuff for a data engineer? I have searched quite a bit but had no idea where I should start from to get a fundamental and good knowledge. Do you guys think courses in Udemy are worth taking?

https://redd.it/zco2fo
@r_devops
Starting a new job tomorrow. First time DevOps Engineer. Any advice?

Went through a boot camp and finally landed a good job after months of bombing interviews. I start tomorrow. It’s a remote job but they want me to come in in person for the first week (company is in my city) before switching back to remote.

I’ve been told the first week would be mostly training and getting set up with the right privileges and what not. Any advice on questions I should be asking or things I should look out for?

Also, what would you expect from a new hire if you were a hiring manager in the first week, first 30 days and first 60 days?

Nervous but excited. Just wanted to see if anyone on here had any advice for me.

Edit: Also, I’m an AWS guy but the company uses Azure, to give some more context. They’ve mentioned that it didn’t matter that I don’t have Azure experience and that they’d train me on it but if there’s something I should be looking out for that’s different from AWS that would be great to know too.

https://redd.it/zcux66
@r_devops
What are frameworks to focus on in 2023?

I’m full stack,

What are some frameworks to concentrate on in 2023?
I’ve tested Vite and also some c/c++ for engineering…
I’m not quite sure on what direction to take,
Frontend…..backend…..something else


What is being concentrated on? Or asked for?

https://redd.it/zcvzk9
@r_devops
How do you manage your Helm values files?

When installing 3rd party charts, do you take the entire values.yaml file and store that in your own repository, or do you create a new values.yaml files for yourself that just contains the options you want to override?

I had been cloning the entire file locally, but as I start to bump the version of some of the charts it is getting hard to manage all the changes? I am wondering if I am better of just keep short values files that only contain my own default overrides.

https://redd.it/zbnnuk
@r_devops
Azure DevOps private agent pool


Hello reddit.

My google-fu seems to be letting me down today. So, I'm working on moving to the cloud and for that to work well, I'm going to need some private pipeline agents within our network. I would prefere if it was container asked, but it seems like docker in docker is insecure and Kubernetes/oprnshift don't support docker in docker anymore (because they moved to containerd and podman)

Anyone of you have any ideas on how to do this in the best/simplest way?

https://redd.it/zd069u
@r_devops
Need help regarding Choosing a Master thesis topic!

llo Everyone, I am a Master's student and I am currently taking my master thesis. My interest is in cloud and devops but i am not very good with it, therefore i wanted to take my master thesis in Devops or cloud, i would appericiate if you could guide me to choose a suitable topic which is interesting and easily doable in 4-5 months i would be thankful for your input.

https://redd.it/zbho0f
@r_devops
AWS Budgets - how long does it take to update?

I'm building a tf setup which will provide sandbox AWS accounts with set budgets and automated nuking / moving to a restricted OU when the budget is @ 100%. Testing some stuff for now, and I'm wondering how long does it take for a budget to update / budget alert to trigger.

This specific info seems (understandably lol) hard to find in the official docs - I presumed it would be 8-12 hours (budgets update interval?), but as a part of my testing I've started up some instances for an hour yesterday, that should've triggered an alarm for my budget treshold, but the budget is still at 0. Maybe I'm misunderstanding something here? Thanks :)

https://redd.it/zd26zt
@r_devops
Urgent help regarding choosing a Master thesis Devops Interesting Research TOPIC; PLEASE HELP

Hello Everyone, I am a Master's student and I am currently taking my master thesis. My interest is in cloud and devops but i am not very good with it, therefore i wanted to take my master thesis in Devops or cloud, i would appericiate if you could guide me to choose a suitable topic which is interesting and easily doable in 4-5 months i would be thankful

https://redd.it/zbhnmo
@r_devops