Using Helm with cert manager

I am working on a creating a new deployment using Helm and would like to know if I have my API GW container running on port 8080 for example how I would go about making the helm chart work with the cert manager? I don't think the port matters but was unsure what I needed to add to the helm chart to make it work with cert manager. I checked the docs and didn't see it. I am sure it's there though and I am likely looking in the wrong place, any chance someone would be able to point me to the right docs or would be able to point me to a guide/blog post about it?

https://redd.it/e6ajl3
@r_devops

In one year, can someone go from help desk to Junior Cloud Engineer?

If they were to learn the following at a fundamental level + with a fundamental understanding of Linux administration and networking:

- Ansible
- Terraform
- Docker
- Kubernetes
- Python

https://redd.it/e6e8q7
@r_devops

Kubernetes implementation for CI

Hey people, I was wondering if you could give some insights on the best/fastest K8s implementation to use that can be run during CI for E2E testing.

I'm in-between either Microk8s, K3s or Minikube.
The viable choice seems to be K3s ATM.

EDIT:
We're using CircleCI.

https://redd.it/e6edme
@r_devops

Managing build dependencies and artifacts

I work on a macOS project which is in a monolithic repo. I'd like to split it up in order to minimise what is rebuilt and to be able to avoid requiring all dependents having to use the latest changes to their dependencies.

It would be awesome but not essential if developer and CI/CD builds could make use of a shared cache of debug and release pre-built binaries, all the way from shared objects to notarised packages.

It has a diverse set of artifacts and a fairly complex build pipeline:
* Static C++ libraries, Qt/C++ daemons and a GUI bundle
* A Swift Finder Sync Extension
* Swift libraries, bundles and a binary
* An Objective C library and a plugin bundle
* An IOKit kernel extension
* Signed and unsigned packages
* Notarisation of released builds
* An installer project which brings most of these artifacts together - the daemons, GUI, kernel extension, etc.

I'm currently investigating Conan and Bazel. We've started to use Swift Package Manager for the Swift/Objective C libraries.

There are lots of ways to do this and I'm sure some will lead to more trouble than they're worth, so I'd like to see if any of you have experience or advice from similar conversions?

https://redd.it/e6f99s
@r_devops

How to get better at obtaining a developer mindset?

Background: I have 10 years experience with infrastructure support/software support and strong Powershell knowledge with some understanding of Python. I’m currently in a role where I’m managing AWS infrastructure / monitoring (Datadog, Grafana, CloudTrail) as well as handling business application batch automation / system checks using an enterprise job scheduler.

My problem is that, while I’m not really considered DevOps within my company, we’ve been slowly ramping up things like source control (Bitbucket / Git for my team), release processes, proper validation testing, and my boss basically told me bluntly that it seems like any time I push something into production it doesn’t work 100% correctly. I’ve never really been in a position where I had to worry about the dev side of things; I’m a complete amateur when it comes to “best practices” of writing code since I’ve never really had my work reviewed and I was just known in my previous jobs as “the kid who could write scripts and get things done” by the older guys.

I know I have a knack for it but I just seem to have trouble hitting the mark with making sure all of my ducks are in a row with testing and releasing to production environments. Anyone have any tips or recommendations to help me improve or change my way of thinking?

https://redd.it/e6b1mn
@r_devops

Need advice on setting up cross account jenkins deployment in AWS

I want to set up a jenkins build from our “Non-Prod” (basically test environment)  AWS account to  another Account called “BI-Prod". I essentially want jenkins in Non-Prod environment to deploy data pipeline in BI-Prod. I am thinking to follow the approach where you a create role in the target account with the correct permission and create some sort of shell script in Non-Prod Jenkins master to STS assume that role in that BI account to deploy data pipeline. Is this right concept to follow ? Is there other way to implement this for best practises ?

https://redd.it/e68hkp
@r_devops

The State of Kubernetes Development Tooling

This is a 30 minute talk from GOTO Berlin 2019 by Ellen Körbes, life-long command-line dweller and developer relations at Garden. Check out a link to the talk along with full talk abstract below:

[https://youtu.be/b1RsNXGLuUk?list=PLEx5khR4g7PKMVeAqZdIHRdOwTM1yktD8](https://youtu.be/b1RsNXGLuUk?list=PLEx5khR4g7PKMVeAqZdIHRdOwTM1yktD8)

With Kubernetes emerging as the de facto standard for service orchestration, the discussion is shifting from "How does Kubernetes work?" to more specific concerns such as "How do I achieve an optimal development workflow?"

The answer lies in finding combinations of tools that work together in synergy. They should produce end-to-end workflows that perform effectively in the real world, while also covering a wide range of development stages—building, deploying, debugging, and so on.

Goal in mind, let's dive into the tooling available in the current landscape—tools such as Helm, Draft, Skaffold, Forge, Telepresence, Garden, and Tilt.

We'll map their capabilities to developer needs, then outline workflows developers can use in practice right away.

**What will the audience learn from this talk?**
Attendees will learn combinations of tools they can use to have an effective workflow for developing applications that run on Kubernetes.

https://redd.it/e6h752
@r_devops

Out-of-state On-Site Interviews

Is it customary for a hiring company to compensate a candidate for travel expenses to/from an on-site interview if its out-of-state?

I've been in the industry for nearly 20 years, but I've never interviewed out-of-state before. I'm looking to see what other people have had for experience. I'm currently operating with a "can't hurt to ask" mentality.

Thanks for any feedback.

https://redd.it/e69ebc
@r_devops

Sonatype Nexus search api does not return all the artifacts

We use sonatype nexus OSS 3.14.0-04 to store all our artifacts and we use Jenkins as our CI/CD took. I introduced snapshot support recently. From Jenkins I use the below groovy script to fetch all artifact versions and populate an Active choice drop-down:

​

`#!/usr/bin/env groovy`

`import groovy.json.JsonSlurper`

​

`#source can be "snapshot" or "release". - source is an Active choice parameter with these two options. I read from the source field.`



`def url = "https://<<<nexus-host>>>/service/rest/v1/search?repository=maven-" + source + "s&`[`maven.groupId=com.xyz`](https://maven.groupId=com.xyz)`&maven.artifactId=rader"`

`def xml = url.toURL().text`

`JsonSlurper parser = new groovy.json.JsonSlurper()`

`Map parsedJson = parser.parseText(xml)`

`def versions = parsedJson.items.version`

&#x200B;

`if (source.equalsIgnoreCase('snapshot')) {`

`Set<String> modVersion = new HashSet<>()`

`for (def item : versions) {`

`modVersion.add(item.split("-").getAt(0) + "-SNAPSHOT")`

`}`

&#x200B;

`return new ArrayList<>(modVersion).sort()`

`} else {`

`return versions.reverse()`

`}`

&#x200B;

The problem is, this does not always return all the artifacts both in snapshot and release. I have to go and delete some older artifacts to make the newer ones appear on the list.

I do not see anything suspicious in the log and not sure how to proceed to troubleshoot this. Is there a better way of achieving the same? Any pointer would be highly appreciated.

https://redd.it/e6c0tb
@r_devops

How did you start out before becoming DevOps?

Hello, I just graduated last year, and currently working as an IT support. I just realize that I have a passion for writing Python automated script to do my boring daily routine tasks. Therefore, I'm thinking of becoming a DevOps in the next year or two.

According to my research, there are not many DevOps junior positions.

* So how did you guys start out before becoming a DevOps Engineer?
* Do you like this job or not? Any hardship?
* What about job security and growth?
* Any advice on becoming a DevOps?

Thank you for reading my post. I'm looking forward to your answers soon.

https://redd.it/e68rfi
@r_devops

Terraform 11 to 12 upgrade

curious on how people handled this upgrade.

We have a large platform/deployment model and it's broken into terraform layers. Different layers interact with each other by reading the state of the layer its interacting with.

It's my understanding that we can't upgrade some layers to 0.12 and except 0.11 layers to also work and vice versa, essentially we need to upgrade ALL of our terraform to 0.12 in one go.

The answer may be as simple as following - https://www.terraform.io/upgrade-guides/0-12.html through each of our layers?

https://redd.it/e62pdv
@r_devops

Terraform support for HCL <-> JSON conversion

I recently found this quote from [a post published by HashiCorp](https://www.hashicorp.com/blog/terraform-0-12-reliable-json-syntax/) back in August 2018:

&#x200B;

>In future versions of Terraform, we will also support native tooling to convert HCL to JSON *and* JSON to HCL cleanly (including comments). This gives additional tooling more power in working with Terraform configurations and also enables people to author configurations in whichever language they feel comfortable.

Does anyone know the status of this feature? There doesn't seem to be any reference to HCL <-> JSON conversion in the Terraform docs, and I'm having trouble finding references to it on GitHub. I've found a few third-party tools that sort of fit the bill (such as [json2hcl](https://github.com/kvz/json2hcl)), but so far nothing that is native to the Terraform CLI.

For reference, I'm working on a project that scaffolds Terraform configurations for various types of infrastructure. Up until now, it was possible to use a template language to generate my TF, however things have become complex enough that I need a programmatic solution.

https://redd.it/e63x41
@r_devops

Api management

I work as a java architect in the computer sector of an insurance company.
As part of its digital transformation strategy, management defined the need to incorporate a product for the management and publication of API (API Management product).
Currently we work with open technologies (JEE / JBoss mainly) and we use SOAP services (both internally between our applications and for external clients that communicate with us).
I've been researching product alternatives but I'm a bit overwhelmed by so much information.
Could you please give me any suggestions?
It can be an already established market product as well as a suite of products or technologies.
Thanks!

https://redd.it/e6mxbx
@r_devops

Thought experiment: You take over 20 servers from a previous admin. There is no monitoring or devops infrastructure setup. What do you do?

Bob Smalls of Small Bobs Inc., a successful Beanie Baby collecting/Historical reenactment company, just fired his previous "IT guy" who administers every server the company uses. You, singular, are reponsible for the entirety of Small Bobs' 20 servers. These servers are of various shape and size and run primarily on Ubuntu and CentOS and they are live in the cloud hosted by an off brand provider.

Of the servers you have something like:

(Linux)

2 mail server

1 proxy

12 application servers

3 web servers

2 file servers

You can of course spin up as many new servers as you want. The only thing you can't do is hire any more people to help you. You also have to take calls occasionally to help end-users when they have trouble with using internal software.

What are the first 5 to 10 things you do to make you managing these servers, and thus your life, way easier?

https://redd.it/e6ng5h
@r_devops

Anyone have one of those days when Kubernetes makes you want to paint the ceiling?

I love K8s but gosh sometimes it just doesn't work as expected.

https://redd.it/e6pj57
@r_devops

To deploy vault on kubernetes or not?

We are an early stage startup on GCP and using GKE.
I want to introduce vault in the ecosystem for storing secrets and I am looking for advice on how should I go forward with it.

To give more details - we have a private GKE cluster with NAT and manage the infrastructure with terraform.

1) Should I deploy vault in the same kubernetes cluster? Or as a separate cluster on compute machines.
2) If same, wouldn't it consume a lot of resources along with consul?
3) If different, what's the best way to authenticate services with it?
4) We don't have any use-case for consul, what would be the best alternative as a store? postgres?

This is my first post here. Any advice is much appreciated. Thanks to the kind ones.

https://redd.it/e6lspv
@r_devops

How can i provision Elastic Beanstalk with Terraform?

I'm setting up an Elastic Beanstalk app to run a Rails API using Terraform.

I need Terraform to ssh in & create, seed the RDS instance on creation. I would imagine i can do this with a provisioner but i'm not sure how & i cannot find any documentation.

The other option i assume would be to use EB container commands through Terraform, but i'm not sure how to do this either.

Any ideas?

https://redd.it/e6pbl5
@r_devops

AppVeyor vs CircleCI vs Travis

So I'm not a devops, but I've spent some time using CircleCI. I liked its features plus the local cli tool was a big plus. However that's the only CI tool I know of.

How do AppVeyor and Travis compare?

Is it just a matter of tastes? Are they pretty much the same? Any benefits choosing one over the other?

https://redd.it/e6ovyj
@r_devops

I get to start a DevOps program from scratch. Suggestions???

Hi all!

I was just put in charge of some pretty big projects at my job. Think creating a data warehouse, IAM, and all scripting support to the NetOps team, to name a few.

Here is the exciting part. I get to create this DevOps environment from scratch. My boss is giving me all the tools I need. I’m already familiar with SDLC/Agile, and that’s the model I’m going with.

I want to set this up correctly so, here’s my ask: what are common mistakes made in DevOps? What are some good resource websites? I saw the YouTube Bootcamp, are there any other good trainings out there? Any suggestions?

Again, my boss is giving me free reign. He’s already pretty impressed with my pre-project reports and trusts me to do whatever’s necessary to create a good environment.

Thanks!

https://redd.it/e6n9rl
@r_devops

Keeping track of infrastructure

I have a dozen of physical hardware laying around being Jenkins slaves. These slaves are actively used by multiple departments for building and testing on real hardware(mostly microcontrollers, but also android/osx phones). \*It's currently a goddamn mess\*. I'm not joking when I say that we don't even know if we have all the machines inventoried and what machine does what because the jenkins labels are all over the place. Some machines have custom scripts running on them to fix unrelated bugs that affect other systems. It's frustrating and needs a big overhaul, without bringing down the current setups. Some of the machines are misconfigured to the point that we can't even access them anymore, others are just unstable.

As part of that we started standardizing jenkins labels, installed monitoring tools MANUALLY(ansible would be nice, if IT allowed us to use it properly lmao i'm in devops this is WIP) and are now basically going to inventory everything.

What kind of tools can be used for this? Currently it's just some random webpage/wiki that needs manual updating, which is incredibly prone to error and a Bad Idea(tm). I need something that can be deployed automatically and monitored real-time. Zabbix is an option but it doesn't seem to have a "public" view without logging in(departments should be able to look at the status, but not touch). Also requires a master/slave setup, and I'd prefer agentless if possible

Other advice is also welcome

Yes, I know normally maintaining infrastructure isn't really part of devops. I just got on the job and have legacy stuff to deal with.

https://redd.it/e6w353
@r_devops

When writing deploy/rollback scripts - pick target based on current branch. Good or bad?

We are having this discussion internally. Some say that is safer to have specific scripts, other that because its automated, having a single on that switch depending on branch is cleaner.

&#x200B;

What are good practices on that? Does that even make sense that we are having this discussions?

https://redd.it/e6wzsi
@r_devops