Auto merge req changes to diferent repo

Hi, I manage several eshops on the Opencart platform where each has its own repository. There are modules installed in them (they are also in the repositories). My question is if when I make changes in these modules can I do a Github action that would create a merge request in the main repository that contains the modules themselves. There I have then created an action that creates an installer from the modules and I could simply transfer the changes via the installer to all opencart installations (if they need it).

So it would be a Github action that would check a specific directory and in case of changes (commits to tracked files) would create a merge request to another repository with those changes.

https://redd.it/z3jkjz
@r_devops

Who else is forced to use Windows and how do you work around it?

I would love to use any Linux OS (MacOS would be okay as well) for my daily job, but I'm forced to use Windows. I would like to compile, test and build software as well as run it locally for test purposes (it sometimes helps to debug a problem without having to deal with the cloud infrastructure first).

I tried WSL but WSL comes with a lot of restrictions and I was wondering how you deal with this? Do you create a VM locally?

Please share your advice :)

https://redd.it/z56dym
@r_devops

HTML and CSS

Hello I am a beginner programmer (still evolving).

Does anyone know good websites for me to copy not too difficult that allow me to put into practice the knowledge of HTML and CSS that I learned ?

Thank You! 💻👨‍💻

https://redd.it/z2z225
@r_devops

NOT ABLE TO DEPLOY LATEST IMAGE IN PRIVATE GKE CLUSTER USING CLOUD BUILD

My gke cluster is private, and I am running cloud build for building an image from dockerfile present in github and pushing it to gcr. Till here, pipeline is working fine but at the end in deployment part it is giving connection denied error. When I am making my cluster public, it is also working.
But not working in private cluster. I have also created private worker pool in Cloud Build by adding

options:
pool:
name: 'projects/$WORKERPOOL_PROJECT_ID/locations/$REGION/workerPools/my-pool'

But still it is giving same error, Pipeline is not running in this worker pool.

https://redd.it/z2mpx5
@r_devops

Replacement for Chef?

Hi. I've used Chef for deployment before and I liked it. One of the things that I liked it that it let me have about 60 servers without paying for anything. Now I have to pay $700 a month? F that. So I'm looking for a replacement to deploy a node.js app to a bunch of Macs, setup some pm2 scripts, things like that. What's the solution that isn't garbage and will take the least time to learn? Thanks!

https://redd.it/z592gn
@r_devops

Cookies - do they understand the content of a website?

Hey!


I'm not a Dev, but I believe some of you guys might actually be able to answer this question.
I'm aware that Cookies are used to track specific actions on a page, e.g., session, to save the shopping cart, chosen language, etc. But the question is - do cookies always identify the website's content to detail?


To give you a bit of context - there are many legal document generators out there, and they rely on generating documents about the webpage based on the info the cookies provide. I have a hypothesis that not always it's precise, as you do not get the full understanding of the page's content just from cookies.


Would I be wrong? Or do cookies usually determine quite precisely the content of the page, e.g. "if there's a form, or if you're asking for e-mail". I'm looking into website analyzing tools and was wondering about cookies. But since i'm not a dev, I am quite still trying to understand the scope :)


Thanks!

https://redd.it/z2mf87
@r_devops

New Relic / Monitoring Tool Alternatives

Right now, we have New Relic as our APM/Monitoring tool, however we are working in the PCI DSS industry and New Relic isnt PCI Compliant.

So I did some research and found out these tools that are PCI compliant:

* Datadog
* Manage Engine OP Manager
* Nagios
* Site24x7
* Solarwind

So I was hoping if any 1 would recommend extra tools, or tell me/ give feedback of which tool do you use (if you used any of the above), or which tool did you walk away from (and why?)

​

And if you are in the PCI Industry which tool do you use and why?

Needs to be on-prem or supports on-prem.

Really appreciate the feedback <3

https://redd.it/z5e1tt
@r_devops

Tekton and ArgoCD Feedback required

My Company is refactoring our entire pipeline and moving away from AWS Code* and we are looking through different CI/CD tools.

Our requirement is to have them on-prem (OpenShift).

We considered Jenkins but ofc, I was heavily against it, considering all the problems it has.

I recommended Tekton for CI and ArgoCD for CD (Doh), due to them being more k8s native and modern.

&#x200B;

I would like to hear your feedback if you worked with any of them(also some tips and tricks) (or if you stopped using them), and what other tools are you currently using?

https://redd.it/z5e67z
@r_devops

Dinamic Infrastructure

Hi,

My company wants that every time a client purchases a service, some new infrastructure has to be automatically deployed (ELB, ASG, IAM, and so on).

To my knowledge just putting Terraform in a CI/CD pipeline achieves this (I created a very small demo using something docker-compose), but I have read that it shouldn't be used for this purpose.

Besides this, I'd like to create an ami pipeline, where the ec2 instances will be created from scratch every time somebody pushes code to his repo, this will be used for the same previous thing, again, is terraform the best approach for this?

https://redd.it/z5kmyi
@r_devops

Oh the interview and the rambling thoughts regarding my life choices in Tech in the moment

Though gainfully employed, for certain masochistic reasons that I can't fully account for, I will take 1-2 interviews every year or so. I set expectations that I'm happy where I am at and that I will go through the process and see if we end up on the same page. I do this to empathize with the people that I interview, to see if I need to update my processes, and I keep an open mind that maybe someone has a unicorn role that will make me happy without taking too massive of a pay cut.

Maybe I'm just utterly apathetic at this point or something else is in the mix, but I felt nothing short of contempt the entire time during my interview yesterday. For the first time ever I thought, maybe I'm in the wrong profession. Not only did i find myself not liking the folks that I would be working with but also questioning what the fuck I've done with my life for the last 10 years.

This shop was a little ass backwards in terms of doing a sys-design first and then the coding challenge and then the behavioral. Mid coding interview I said, I think I'm good, I don't really want to do this anymore, said bye and called it a day.

&#x200B;

I'm at a MAANG company. Pay is great, benefits are great, but I'm starting to realize that this life is not making me happy. I left Dev because I thought I would find more work/life balance and I'm as miserable or more than every.

https://redd.it/z24s5j
@r_devops

How do you manage your SSH keys?

I’ve heard different schools of thoughts. On one side, you can generate a new SSH key on each computer and copy to your servers with ssh-copy-id

The benefit I’m told is that if a key gets compromised just revoke that key everywhere and you’re good. That’s great but if you work from a dozen computers and servers, knowing which key to delete could be a challenge

On the other hand, I could generate a single key pair is sneaker net (or scp it) it to each computer I work from. Then, if your key is compromised, you deleye that, generate a new key pair and keep going.

I’m not so much in Devops but I am an enthusiast, so I’d love to hear what everyone else does.

https://redd.it/z5nv6w
@r_devops

File from jenkins needs to migrate for postgresql

I have a job in Jenkins which generates .csv file, I need to use that .csv file in postgresql. How to do using jenkins pipeline?

https://redd.it/z258so
@r_devops

can i get a devops engineer job as an entry level?



Hello, I am a junior majoring in cs. . I am interested in devops. However im not sure if i can get an entry level job as devops. Can anyone tell me if its possible? if then what course would be the best?

https://redd.it/z23l71
@r_devops

How do you deal with bad commits on live production environment?

Most of the pipelines are in place, but the issue is with the general infrastructure of the project. What I'm talking about is our main "product" - An application built out of 5 ECS clusters. We have 3 stages before prod, but we had an instance where prod failed due to an AWS account specific issue we oversaw, leading to some downtime on our most important, and only customer facing environment (Terraform did half a deployment due to a missing secret, which meant I had to quickly add it and run the pipeline again). I'm just wondering, how would you approach this issue? I understand the concept of blue/green deployments, I'm just trying to figure out the best way to apply it here. Add another (test) domain on prod and have it deploy there first, before reaching actual customers? Or have terraform/pipeline instantly revert to last commit if any issues are found?

https://redd.it/z1vn9a
@r_devops

looking for a mentor

Hey there,i'm staring my journey in the field of devops i need a mentor or a buddy who can help me with my journey. Thank you

https://redd.it/z1sn52
@r_devops

Multi Organisation Architecture.



Can someone help me find some resource, I have no idea what its called but the concept goes as follow

So, I want to build something where if an organization registers on my site they will get their own domain. Let's say for org ABC it will be abc.myapp.com similar to how JIRA does it.

Now I am not sure if they create a dedicated database for each organization or use the same database and have some kind of relation. Is it possible to spin up a new database for the said organization? If yes, can anyone provide me the resource or terminology of what it's called?

Let's say I write API to handle all the necessary things for a single org. Is it possible to create a whole server using my code for each organization? taking the above example create-task for ABC organization will route to the server of abc.myapp.com/api/create-task. So every organization gets its own server and dedicated database with all the respective databases.

Also If I update a feature or change the UI then all organizations should receive the new updates on successful deployment.

Can someone please help me find out the DevOps tools require to do such automation?

Thanks, any help would be appreciated.

https://redd.it/z5vqb3
@r_devops

How to enable network policy in running AKS cluster?

I want to implement network policies in my AKS cluster, but the cluster was deployed without network policy before my time.

I can see network policy as ‘None’ in the cluster network config page on Azure.

Is there any other way to enable it without recreating the cluster?

https://redd.it/z1patz
@r_devops

How to create a server status service

I'm interested in knowing how github and other websites create things like githubstatus.

I'm trying to create something similar for my website. Is there an open source project that provides something like github status.

https://redd.it/z61hxh
@r_devops

Should I make my resume simple or more detailed with only 1 year of experience?

Here are some contexts, I'll make it as short as possible. So I will list some overviews of what I've done so far by quarter.
I was a self-taught engineer(Spanish Undergrad), so at first, I was hired as "assistant engineer" and only one of backend developer partially handles deployments , before this job I had "no experience in any of the tools or knowledge but only Javascript", our "dev" environment consists of services and databases running on bare-metal machines using docker-compose and few of them on local Kubernetes cluster, the deployment pipeline was done by triggering Jenkins slave running bash scripts on target machines.

Once service is tested by project managers, it is then shipped to UAT on GKE, as DevOps team will handle the rest.

Q1:

1. Take over full responsibility of managing existed deployment pipeline with optimization and centralized configs over different environments
2. Fully responsible for "local development infrastructures" including handling system or service failures, debug along with the dev when something unusual happens from the log
3. Test, research infrastructure solutions and configurations like database failover, or monitoring systems
4. Handling creation of Dockerfile based on various request

Q2:

at Q2 because I've basically handled tasks that the team leader expected and during this phase our team was super ahead of the schedule, so he didn't assign new tasks to me, with this much free time, most of the things I did in the office were fairly irrelevant to my job like playing around Neovim or just straight up studying documentations

1. Created a four-stage integration testing pipeline along with Slack alerts
2. Write tests for two of the services using Javascript and created a small library for future test in these two.
3. Spent almost 3 weeks to figure out how to apply progress bar to the test, only to find out it couldn't be displayed because the test log is displayed on Jenkins console output. (but hey I know ANSI Escape now)

Q3:

1. Take over responsibility of " frontend team's deployment pipeline" and optimize them with Jenkins pipeline using groovy script to build parallelly and save more than 50% of the time
2. Created a Gitlab CI pipeline and runner for linting existed Golang code base and also linted two of the service with golangci configuration along with teaching some of the devs how to set up linter in IDEs.
3. Fully handle all the infrastructure and services for "Dev" environment meaning if devs found their build failed or want to add new services/requests, I would handle that.

At this point as I found myself doing "less assistant job", I transferred to DevOps team since they were recruiting at the moment, as I found great interest in Kubernetes, networks and Linux systems.

UAT and Prod environment uses totally different flow, with GKE along with GitOps tools, again, I had no prior knowledge of these things before except for Kubernetes because I tried set up local ones with kubeadm but still at a very entry level lol, and since DevOps team has only 3 people including me atm, they handle every product in the company, I had to learn lots of things quickly and be able to pick up some of the tasks.

So, I studied and worked my ass off to a level where I can at least identify the problem and tweak things comfortably while understanding what I'm doing.

Q4: (current time)

1. Maintaining existed production and UAT GKE cluster and deployment, upgrade some of the infrastructures and migrates GitOps tool
2. Migrating one of the main products from VM to GKE(also UAT and prod), set up infrastructures like nginx-ingress or prometheus and services using ArgoCD and helm charts, also handles deployment pipeline and network policies like gRPC load balancing
3. Set up blackbox exporter inside GKE and wrote a small in-cluster operator with Golang that syncs service-monitors according to the data from google sheets (with client-go)
4. Test and research some stuff like SSO solution with Okta or

SAML2, or streaming logs from CloudRun to Elasticsearch via logstash etc.
5. Fixed many bugs for a tool written in Golang

As now I'm heading to Q5, and I'm planning to reach out to new junior DevOps opportunities since it looks like there's not much space to negotiate salary with my company as they're still providing me assistant level salary.

After listing things I did, I feel like I've done many things that can potentially be on my resume, as I've seen many people recommending resume should be as simple as possible, I'm quite worried if I put an overview on my resume, it would seem like I'm bluffing on my resume based on my experience and background, though in reality I spend a lot of my free time studying, and if I put too much information, they won't be interested at all.

thanks for making this far, what do you guys suggest?

https://redd.it/z62bl5
@r_devops

Help me understand real use cases of k8s, I can’t wrap my head around it

So from what I’ve read k8s is for mission critical dockers you want to provide high availability for or scale up. Correct me if I’m wrong!

After running dockers 24/7 for years I’ve never had a container randomly fail or been overcome with too many connections to the point where I’m thinking “if I had more this would have solved the problem “. So in terms of high availability I don’t get it. From what I understand k8s does not even sync data between nodes since they’re all using the same volume mount which to me, is the complete opposite of high availability. To me intuitively, k8s should be something that literally syncs multiple containers all with their own individual volume mounts across multiple remote locations.

In terms of scaling, at what point is a load balancer just not cutting it for you anymore? Such that adding more nodes is the solution.

Who actually benefits from k8s? I see too many examples of enthusiasts deploying at home because they can instead of actually needing it, and when I ask for production examples the only thing i hear are examples of Google, the biggest tech company on earth.

I really am not trying to attack k8s and would love to deploy it myself if I see a real benefit from it.

https://redd.it/z64b1q
@r_devops