Should I be looking for non-junior roles now?

Hey all!

Currently I graduated a year and a half ago and was very fortunate enough to jump straight into a Junior DevOps role and I've been there since. However, the company I currently work for is moving to another city (In May), and I'm adamant on staying where I am for various other reasons. Now when I'm applying for new roles in May, I would have accumulated just shy of 2 years experience as a DevOps Engineer.

Is it naive/complacent of me to stop looking and applying at Junior roles and apply for normal DevOps engineer positions? Or should I strictly be applying to junior position until a certain point etc.

The only reason I ask is that I want to be sure going through the whole, learning everything from scratch and being taught everything from ground up again (like my current did for me) is the best thing to do in my position.

NB: I know judging experience by years working is not an accurate indication of how good someone is, but the question is just for some general advice for those of you who have been in the industry far longer than I have.

​

Thanks!

https://redd.it/egtumq
@r_devops

Cors issue

I have run into a cors issue that doesn't seem to want to work, I have a Vue.js application which connections to the backend (node.js), front end is \`app.example.com\` and the backend is \`api.example.com\`

The backend has cors enabled on the api gateway which works fine in testing locally which leads my to believe it's an issue with the kubernetes deployment or maybe aws setup.

I am using Kubernetes with the nginx ingress controller to host the api with the front end on s3 as a static site. I have this anotation added to my ingress controller though I added it after creation.

nginx.ingress.kubernetes.io/cors-allow-origin: "https://app.example.com"

​

My ingress controller is an AWS ELB which is created by my ingress resource.

​

Anyone have a suggestion on what I can try? I've done a good bit of google time and came up short past enable cors on the backend (did before issue), enable cors on the bucket (not sure it's needed since I have no issue connecting to the front end just sending to the backend) as well as adding that anotation to the ingress controller and applying that

https://redd.it/egur1f
@r_devops

Thinking of starting a new project - Cloudformation Governance

Hi,

Thinking of starting a new project to plug into a CI pipeline for CloudFormation. Thinking of putting together functionality to parse an incoming cloudformation script to enfore governance rules i.e [0.0.0.0/0](https://0.0.0.0/0) ingress, port 21, tags missing etc. Closest anology (and inspiration) is Sentinel Framework from Hashicorp.

First off, useful?

Second, trading off simplicity of implementation and use vs. flexibility - go for a command line utility with pre-defined rules (but designed in a way to make it \*relatively\* easy to add rules) vs something more like Sentinel where it is all written as psuedocode?

Tom

https://redd.it/egtai9
@r_devops

Making my own PaaS (some thoughts and questions)

First of all, hello and happy holidays.

I'm not much of a system admin or a devops, I'm a developer who knows a little about how devops stuff works. Because I did deploy a lot of projects I've done. Thanks to my friend who runs a PaaS startup and thanks to docker! But I was thinking about starting a PaaS startup myself. I thought I pay much less (compared to the time I had to rent a VPS and probably hire an admin to deploy the damn thing) and I earned much more (less deployment time and money meant more projects of course) and that's too freaking economic.

So, I did a little search and found Dokku. It seems this piece of software is very minimal. I am looking for something similar, but with a proper accounting system or something, then I can give people services they need, just like my friend.

Is there anything available?

P.S : I'm not asking him for the tool, because I did it once. They developed the tool themselves and they don't share the codes.

https://redd.it/ege12s
@r_devops

Critique my Terraform setup for a single tenant application.

We run a single tenant application in AWS, where for each client we sign on, they get there own set of infrastructure.

On the AWS side, i'm planning on using an organization & having an account for each client.

Each clients infrastructure should be nearly identical, minus configuration differences, like auto scaling group size for example.

What i'm planning on doing, is

- having a set of versioned, generic modules for each piece of the infrastructure, ie database, backend app, networking, security groups
- having a set of live, environment modules that use these generic modules to create a set of infrastructure & set specific configuration.

Specifically for production, i need to have a set of infrastructure per client. So to avoid code duplication & ease updating the modules, i am going to create a base production module, that imports all of the generic modules. Then a module for each client that uses this base production module & configures the remote backend, sets all of the specific variables & so forth. The idea is this should solve the issue of having to update a ton of client production modules when changing infrastructure.

Thoughts? Does anyone see any possible issues with this?

https://redd.it/egdcwm
@r_devops

AWS autoscaling help.

If I have a server running heavy application server like WEBSPHERE APPLICATION SERVER or WEBLOGIC server will AWS be able to create new instance and distribute the load to newly launched server? Installation of light weight applications like Jboss or tomcat is easy but not WAS etc. that’s my dilemma. Thank you.

https://redd.it/egb5ln
@r_devops

Trouble recording videos

We have cameras which stream video. And for every user session, we record the video.
I have a video streaming service (using node-media-server), onto which the camera streams the video all the time. And there is another recording service. Whenever a user performs an auth operation (logs in), I spawn a process from the recording service, and create a write stream. When the user logs out, I kill the spawned process in which recording was happening , and I upload the video to google storage bucket.


My problem is with videos of 0 bytes. It seems be happening thrice out of 50 times, on average.

The problem starts when the pod restarts. I just have a single pod. ( which is enough for my cpu and memory requirements. I just have 20 cameras currently from which I stream) . Also, there's this constraint, that if I were to have 2 pods one on VM1, and the other on VM2, then I would need to know which VM (or pod) the process will be allocated. I think in the very near future, I will have to increase the number of pods, and I'll face this problem very soon.


Whenever the pod restarts ( due to reasons that are still unknown to me. I have checked the containers audit logs to find out why my pod restarts, but that didnt help me much, also the resources I have requested are sufficient for my load) and if there are any processes in which recording is actively happening, they will be lost. And my recordings fail. I believe this is the reason why I see those 0 bytes videos.


How do I ensure that if at all the pod restarts, the active processes should not immediately be killed. Or maybe, if there's a way to delay the pod restart until the current recording finishes ( i highly doubt if this is possible tho).

https://redd.it/ega99j
@r_devops

What's your DevOps elevator pitch to non-tech family and friends this holiday season?

Title. Family Christmas party tonight so I have to be prepared.

https://redd.it/ef8g5m
@r_devops

Hosting Wordpress On Azure (Step by Step Tutorial)

In this [tutorial](https://youtu.be/api6HM4kpWc) we will deploy a WordPress blog on to an Azure App Service for Linux. Though there are many ways to deploy WordPress on Azure like deploying bitnami WordPress and WordPress for azure, we will go on traditional web deployment approach to have full control over the app resources on Azure.

Steps:
1 Create App Service (Linux) 00:19
2 Create MySQL Database Server 03:14
3 Access Secure Shell (SSH) 07:30
4 Deploy WordPress 09:20
cd /home/site/wwwroot
wget -c https://WordPress.org/latest.tar.gz
tar -xzvf latest.tar.gz
mv WordPress/* /home/site/wwwroot/
rm -rf WordPress
rm -rf latest.tar.gz
5 Connect Azure MySQL Database from MySQL WorkBench 10:15
6 Create a Database for WordPress Application 13:02
7 Configure WordPress Installation 13:42
8 Scale Up/Down App Service Plan 18:47
9 Configure SSL for WordPress on Azure (Important Step) 19:53
9.1 Enable HTTPS only Protocol On
App Service -⫸ Settings -⫸ TLS/SSL Settings -⫸ Under Protocol Settings -⫸ Switch HTTPS only flag 'On'
9.2 Install Really Simple SSL Plugin for WordPress
In WordPress admin dashboard, go to add plugins and install Really Simple SSL plugin
In Really Simple SSL Configuration click on Go ahead, activate SSL
In settings Enable WordPress 301 redirection to SSL

https://redd.it/ef62m2
@r_devops

Handling canary environment in k8s

Hey all,
How do you handle a canary environments in k8s?
I tried nginx-ingress(which I already use it) with it's canary option but I couldn't configure sticky session as well, I saw many issues on this.
So, how you do it?
I'm looking for a solution that will not require any code manipulation if that's possible.
I saw Istio but this is such a large solution for my simple problem.

https://redd.it/ef5a44
@r_devops

Devops Career suggestion needed !

Hello everyone
I want build my career in devops and currently a job based on Linux and monitoring is offered to me so should i go for it for the sake of experience so that i will no longer be a fresher
Because earlier i have applied for multiple devops position and got rejected because i had no experience in the field
( I'm RHCSA,RHCE, OPENSHIFT, ANSIBEL certified )
Or
Should i wait for right opportunity ?

https://redd.it/ef4sff
@r_devops

[Question] Authentication for third party applications behind a reverse proxy

Let's suppose that you have third party web applications, like Prometheus, that do not provide an authentication system. and you would like to have a dynamic authentication to access those resources behind a reverse proxy.

Nginx only supports static authentication from a password file (HAProxy does too). One solution could be to use an authentication system connected to your external different systems like OpenLDAP, JumpCloud,... and allows you to authenticate as described in \[this article from Nginx's Blog\]( [https://www.nginx.com/blog/nginx-plus-authenticate-users#ldap-auth-flow](https://www.nginx.com/blog/nginx-plus-authenticate-users#ldap-auth-flow) ).

​

Do you know tools that could do that? Is there a better way?

https://redd.it/ef4niv
@r_devops

How tech is devops?

I keep hearing that devops is a mindset and a culture but in practice it seems to be almost exclusively about tooling and coding and the application or approach to these in an end-to-end environment. (Just look at most of the content in this sub).

I’m an Agile Coach and I’m trying to skill up. I have a background (a long time ago) in C++. The devops-as-a-mindset thing, to me sounds like hijacking .. ie that mindset is actually no different to an Agile mindset. Is there a difference?

In terms of the technical side, I’m wondering how technical I have to become to sell myself as a devops practitioner or coach, and where to start.

I have no inclination to become a technician again but need a deep enough understanding to play a valuable role in that space.

Does it even make sense to “coach” devops though? If the cultural aspect is really just Agile (albeit in teams that span dev and ops) and the rest of it is basically tooling, then that latter part of it sounds like it’d be more about training than coaching...?

So... a few questions buried in there. Thoughts? What dy’all think of Agile Coaching in a modern devops environment? What advice would you have for an Agile Coach looking to skill up?

https://redd.it/ef1oo9
@r_devops

Credentials not set correctly terraform

I've just got into using terraform and ansible to deploy some cloud infra. I have an issue where regardless of the provider is use ( just testing so far) i get

>`Error: credentials are not set correctly`
>
>
>
> `on <empty> line 0:`
>
> `(source code not available)`

I'm using Terraform v0.12.18 on Ubuntu 1804

I've tried declaring my providers in a [provider.tf](https://provider.tf) file and variables in a [variables.tf](https://variables.tf) file

>`variable "linode_api" {`
`default = "TOKEN"`

`}`
>
>`provider "linode" {`
`token = "${var.linode_api}"`

`}`

&#x200B;

i've even tried hard coding the provider in the [main.tf](https://main.tf) with the token in the file. I know the above is bad security practice however it also throws the same error if use ENV variables

https://redd.it/ef3o89
@r_devops

Automatically list all your services (Zero Conf)

Hey guys, I made a dashboard with services autodiscovery! Might be useful if you use Traefik.

I hope you will find it useful:

https://ms-jpq.github.io/simple-traefik-dash/

The idea is that since Traefik already knows all about your egress, then why doesn't it list all the ones you can reach deterministically?

Simple Traefik Dash basically parses Traefik's routing syntax from its own API endpoint, and assembles a webpage from those.

You can also customize it a bit, but its all in all pretty simple.

https://redd.it/ef3au6
@r_devops

A Comparative Analysis of Kubernetes Deployment Tools: Kubespray, kops, and conjure-up

This 12-page report compares Kubespray, kops, and conjure-up across installation simplicity, configurability, scalability and distribution, supported cluster operations, performance, and debugging simplicity.

Download here >> [https://www.altoros.com/research-papers/a-comparative-analysis-of-kubernetes-deployment-tools-kubespray-kops-and-conjure-up/](https://www.altoros.com/research-papers/a-comparative-analysis-of-kubernetes-deployment-tools-kubespray-kops-and-conjure-up/)

https://redd.it/eeyhmg
@r_devops

Creating a log management system as College project.

I am in last year of my computer science engineering. There are different hackathon are taking place in my college, one of them is sponsored by different government companies where we have to choose and provide solution for there problems. I did selected a problem statement which involves designing a system for log management.
Bellow is the problem statement and description provided be the company:-

--------------------------------------------------------------
Problem statement:-
Creating Ability to Server Logs Management for Analysis & Report of Heterogeneous Servers .

Description:
Servers are the crux of our IT infrastructure and are often exposed to great risk. The servers are covered under the layer of firewall, but they are still vulnerable and require security measures. The major concern is the source of threat. Extensive usage of remote connectivity causes improper security and lack of monitoring measures on remote networking equipments. It is required to keep an eye on the wrong doers, the data thieves and their work pattern. BEML has about 30 servers housed in Datacentre in heterogeneous environment with OS HP-UX, Windows based servers. Outcome expected: Requirement of Server based Monitoring & Analysis Tools to alert the servers with Dashboard report on 30 heterogeneous servers available at our Datacentre.
-----------------------------------------------------------------

I did research and found solution like ELK stack, Fluentd, Greylog. Should I go with these solutions and create something different.
I am good at Linux(I am a Redhat Certified System Administrator), Python and AWS Cloud.

As different students across the country are participating in the hackathon there will be many who are working the the same problem statement as of mine, Please suggest me how I can design a efficient system which helps me to get myself selected within many.

Questions that are arising in my mind:-
1. Why would they choose my solution, if I am not creating something new and also I can't create something like ELK from scratch.
2. What if I choose ELK( let suppose) and some other student also chooses the same? ( As I realy don't have any idea how many students are working on the same project).

Please help me to design a solution for this problem statement.
Thanks

https://redd.it/ef14yk
@r_devops

lamp stack architecture

is it good practice to insall lamp stack on one machine? server / database / cache

or better use something like amazon RDS?

https://redd.it/eezfjh
@r_devops

What should i do to consider my self a DevOps engineer? Since i have no certificate yet

Hello Guys, i was a Android Developer/Java some years ago and in the last two years i'm into this DevOps world, so i managed some tools, i have knowledge at programming, deploying, testing, continuous integration, continuous delivery and some other tools.

But for some reason i'm a scared to say that i'm a devops engineer right now.

What kind of questions would you ask to define that i'm a DevOps engineer?

Some stacks/techs/programming languages i managed in a not so far time:

Android, Java, Flutter, Javascript, Nodejs, serveless, pwa, vue, shell scripts, groovy and some others

Some tools i managed in the last two years:

Jenkins, Gitlab CI, AzureDevops (recentely), ansible, packer, terraform, docker, grafana, CI concepts, CD concepts, AWS (ec2, groups, vpcs, amis and some other tools), sonarqube...


What questions would you ask to me to know if i'm a real DevOps engineer?


I know that certificates are cool, but what would you ask to someone with a cerificate?

https://redd.it/eewe4w
@r_devops

Recommended books for microservice architecture?

Hi all,

I'm looking for book/resource recommendations for architecting with microservices.

Given some set of functionality that my service needs to achieve, how can I best decompose it into microservices? What qualifies as a good design vs a bad design?

Design patterns for microservices are also of interest!

https://redd.it/eey3zg
@r_devops