What are you using for handling incidents in the cloud?

How well is it working?

https://redd.it/q35u6k
@r_devops

Post monolith success stories?

A little groggy after another PM deploy and I'm feeling discouraged about the state of things and don't have the expertise currently to pave the way forward.

On my most recent gig at a small, not new company, we've been trying to make some headway breaking down the mono-repo and it's not been encouraging. Devs are stuck in the way they've been doing stuff since mid 2000s and like most of us aren't super into change. So cultural and technical friction on all fronts, though in conversations people are into it, but afraid of breaking shit.

We're currently hosting everything on VSphere and are tied into Oracle. The approach to ween them off of Oracle is anything but clear. I don't at hand have any good ways to suggest how we deal with apps having different schema dependencies or how we can gradually shift our architecture towards intermediary steps at building and deploying microservices.

I've never worked at a shop with such stringent coupling and dependencies, nor have I had to concern myself with architectural decisions way out of my comfort zone (seemingly everyone else at the company as well) to try and address this.

For the sake of a better work life balance I feel like my org should be making this a number one priority to keep Devs from late night deployments. Additionally, it's just clear that the ripple effect and coupling is beyond what anyone can keep track of in their heads.

Currently reading The Tao of Microservices which has been a gold mine, though I'm 1/3 of the way through it. Hoping the sections on dealing with monoliths is helpful.

One of the founders of Acloud Guru has a video describing their evolution and how they've always been serverless. He noted how they managed to pickup more efficiency and decouple more and more over time. One decision in particular about how they found ways to stand up environments with less and less dependencies using ETL jobs so that a CACHE/DB etc wasn't necessary for all environments really rang true.

https://redd.it/q2kpgp
@r_devops

Using Jenkins for development server deployments

Hi guys, I just started working as a Junior DevOps and I'm looking to automate some deployments for my software development team.


They have a Plesk development server where websites are stored and accessed via subdomains (webpagex.ourdomain.com) before going to production. My question for you is:


Would it make any trouble if I installed Jenkins on that server and run the CI/CD together with bash commands without having to SSH into the dev server?


FYI, I have already created a docker service for Jenkins and its accessible via jenkins.domain.com. If you would suggest a better practice, I would be open to follow. Thank you.

https://redd.it/q2k191
@r_devops

Need help chosing the next work laptop

Hey people, I'm working in a tech company where 99% of tech employees use Intel MacBooks, and it's time for me to get a new laptop.

Right now I'm working on 13inch MacBook Pro from 2018 and I've grown to hate it for a few reasons:
- battery sucks
- if you turn on docker, battery becomes even worse
- heats up a lot
- if you turn on docker, it heats up even worse
- butterfly keeyboard (luckily docker doesn't affect this)

As a replacement i see myself going in one of three directions:
- another Intel Mac (at least keyboard will work properly, but still, this is the choice I'd like to avoid)
- M1 Mac
- Thinkpad X1 carbon gen 9 (running Linux)

I'm using Thinkpad X1 Carbon gen 6 as my personal computer and I'm very happy with it - thus the idea of running Thinkpad for work as well. My concern is that I'll be the only guy running Linux in the whole team. I can't think of a concrete problem with that since we are using Unix-based tools and are running our code on Linux machines, but still i guess there is advantage of everyone on the team running the same OS(X).

On the other hand, I've heard good things about M1 Macs, but the two biggest gripes with it are:
- not sure how well docker runs on it, but it's hard to imagine it will be close to how it runs on a Linux machine
- it might be using the same-named OS as the rest of the team, but it's a CPU with a different architecture, i can imagine that this can result in some issues.

So tldr of the question is:

Should i be the only guy on the team running Linux, the only guy on the team running an ARM CPU, or should I just get another shitty Intel Mac?

I guess the question doesn't make much sense without some perspective of what tools I use for my day-to-day work:
- a lot of docker
- a lot of terraform
- a lot of kubernetes
- some bash, python, go and js
- a little ansible

Not sure how well i managed to formulate my thoughts into a question - sorry if I did a poor job! Any insight would be appreciated :)

https://redd.it/q399jj
@r_devops

Battle-tested tricks for efficient CircleCI workflows.

Our super talented Simon shares tricks from his toolbelt:
https://dev.to/otomato_io/circleci-the-matrix-3c17

https://redd.it/q2qjs5
@r_devops

What means "Always ship trunk"?

I'm having a tough time finding a good working definition or meaning really for: "Always ship trunk". I realize this is an older mantra in one of Allspaw's prezzos but I'm kind of a devops noob. I would like to understand what it means and any other comments or discussion on it is a bonus.

https://redd.it/q3bzaw
@r_devops

Your views on daily stand ups

I’ll start with mine and I accept it might be very controversial…. Doing them everyday is a total waste of time especially if they’re run badly like ours often our (to the extent where our scrum master asks for updates on tickets in the sprint backlog). Doing them every other day, fair enough. If there’s anything hyper urgent it can go on an email and a dedicated session can be set up to discuss it in detail (which should happen anyway). If other team members need help, they should feel comfortable going to anyone at anytime and asking even if the response is “I’m busy now, let’s talk later” or something along those lines.

I’m interested to know what you all think of daily stand ups! Maybe I’m just anti them because ours are so awful!

https://redd.it/q3e1ze
@r_devops

Devops hackathom ideas

Hi folks!
Hope everyones having an amazing day :]

Had a question to all the devops gurus out there. Im very very new to the devops world but would love to join a hackathon that's coming up....
What do you consider a good devops hackathon project? are there any particular problem statements that we can use devops to solve efficiently?

https://redd.it/q3elfv
@r_devops

has anyone here created Azure DevOps using TF? I would like to get your experiences and or pros and cons.

TF seems limited on the configurations one can make in Azure DevOps. But that could be by design

https://redd.it/q3i0g0
@r_devops

Programming language for DevOps

Hi mates, I need to improve my programming skills of course not need not high level. But I can't decide which language is better for DevOps. I know Go is popular and a lot of important tools were created with Go. But I couldn't like Go it is scaring me. I like python and iI have Python basics, so python may be easy for me. Help me about deciding.

P.S I know Bash is better :))

https://redd.it/q3hx7m
@r_devops

First day on the job

What are the things you would do on your first few days as a devops engineer and whom in the organisation would you contact and what sort of questions would you pose to different leaders?

https://redd.it/q3islu
@r_devops

Top open source CI/CD tools for Kubernetes

https://cloud.redhat.com/blog/top-open-source-ci/cd-tools-for-kubernetes-to-know

https://redd.it/q3mmiq
@r_devops

Best way for release deployment of on prem/self hosted Python Flask apps and api's?

A bit of background. My team runs a self hosted Gitlab community edition server and have several Flask based API's and services that we manage deployment, configuration and testing for.

Right now I have Gitlab CI setup to perform linting, sonarqube analysis, unit testing and artifact posting. When all is said and done and testing looks good my team will run Ansible playbooks with environment specific variables either from Tower or CLI to call roles to redeploy, configure and start whatever application or API was updated.

I am trying to move the last piece, the deployment piece, to Gitlab CI. I have looked into possibly calling Ansible Tower's API to launch a job to perform the deployment of the respective project/repo/branch. The other idea I had was creating a docker container with the plays and roles needed for the deployment, pass in environment specific variables and run that in a deployment stage called out in my .gitlab-ci.yml file.

I am curious to hear other ideas or solutions on

https://redd.it/q3oy8r
@r_devops

New Job, new opportunities

So, I will be starting a new job soon, it will be to manage a large test environment. My guess is that it is a mix of Windows and Linux. I didn’t get much detail as the high level managers were not too technical. I know they have been running a devops culture for a few years now. Terraform, Git, etc are things that I can recall. I come from traditional Infrastructure.

How can I prepare myself for this new world?

Thanks.

https://redd.it/q3nrx8
@r_devops

Is kube-proxy an actual proxy or does it just handle networking?

When I first leaned about kubernetes components, I assumed that kube-proxy was a network or http proxy like envoy. However I recently read an article that says that it just modifies the iptables on the node and doesn’t do any proxying.

https://redd.it/q3mgzd
@r_devops

Orchestrator on top of GitOps/Infra-as-code solutions?

I'm considering how to do a higher-level orchestration (and possibly self-service portal) on top of our infrastructure. We have several different technology specific domains where we have different automation and management systems, and of course lots of support systems like inventory, IPAM etc. Building something to make use of the inventory, IPAM system and so on seems pretty straight forward via REST APIs, but I'm struggling for what's the best solution for the technology domains where we employ more of a "gitops" approach.

Basically we have a git repository where technicians would go in and change some YAML files and then commit/push to a central repository and from there the automation system for that domain would go in and configure things according to the definitions in the YAML files. How would you go about editing YAML files in git repositories in an automated way? Have you seen any projects doing this, some code for how they manage it?

Specifically I'm thinking what do you do if some human commits changes to the git repository after the orchestrator or self-service portal read the files, but before the orchestrator commited files. You will end up with some merge conflict situation which seems very messy

https://redd.it/q3tc5c
@r_devops

Best Practices for CKA Exam

removed

https://redd.it/q3tv7j
@r_devops

Elasticsearch Kibana

Hi all. I need to add alerts to my dashboards. I’m using elastic helm chart. Only stateful set is deployed. How can I use config map and configure alerts in elasticsearch.yml.Need your help. Thank You.

https://redd.it/q3t63j
@r_devops

Is there some evolution possible after being DevOps engineer?

Monday I'll start a DevOps junior job, I was wondering what is the common next step in a career after being DevOps engineer ?

Thanks

https://redd.it/q3wj46
@r_devops

What SREs Can Learn from Facebook’s Largest Outage

Likely the most analyzed outage this year given we were all frantically refreshing Instagram (at least I was).

There were some good lessons for SREs we wrote about here but curious what (if any) was a takeaway for you!

https://redd.it/q3yvc4
@r_devops

Should I get into DevOps for money despite being interested in web development?

I am an accountancy graduate but gonna do my Masters in Computer Application. I want to enter IT field so I am deciding which role should I give all my focus to. My Masters would be of 2 years, so I wanna give my all to learn about the role I am gonna choose. As the title says, I am inclined towards DevOps because of the Stack Overflow Survey which showed it being one of the highest paying role. Although, I think I am more interested in learning web or android development.

So should I get into DevOps just because it shows better financial growth? I think I can do either of them, but I have done few personal projects in web development so it would be easy for me since I have little to no knowledge about DevOps. I am also learning Python.

https://redd.it/q3v72w
@r_devops