Provisioning a highly available load balancer in Hetzner Cloud with Ansible, haproxy and keepalived

Hi! For anyone interested, German provider Hetzner has published an article of mine in their tutorials library on provisioning HA LBs with Ansible. Some bits are specific to Hetzner Cloud ma can be easily adapted to other providers who have APIs to manage floating IPs for the failover. Here's the link, hope it's useful.

https://community.hetzner.com/tutorials/howto-highly-available-load-balancer-hetzner-cloud-ansible

https://redd.it/g2iek3
@r_devops

Testing & validating AMIs

I'm building AMIs with Packer right now, and I'm interested in writing tests to validate my images and make assertions about their state, to basically automatically verify that all the dependencies I need are properly installed.

I could just write a shell script to do this, but I'd like to avoid reinventing the wheel if possible. Are there any tools/frameworks that can do something like this?

https://redd.it/g2mhoo
@r_devops

Where can I go constructively volunteer my free time and continue learning while unemployed?

I recently got terminated from a position and, due to the pandemic, worried it might be a little while before I can land the next position. Anyone know of good volunteer opportunities with non-profits out there where I can go volunteer my time and continue working on DevOps projects? Just want to make sure I'm not staying idle and find a way to continue contributing.

https://redd.it/g2hswf
@r_devops

Does anyone here have a side devops gig?

What sites/resources do you guys use to pick up short term contracting jobs in devops field. I live in Alaska, so most of my options are most likely remote jobs anyway.

Apologies in advance if this isn't the right sub for this.

https://redd.it/g2m2mr
@r_devops

SSD OR HDD

Hey folks, I'm trying to run a multithreaded python application on EC2 instance, which type of ebs storage should I use? The machine is basically for migration so lot of memory, network and disk resources will be utilised. The data could be anywhere around 50 TB, do you also think if amazon efs would be useful, I'm a newbie in AWS...

Thanks!!

https://redd.it/g2i0pw
@r_devops

how to land job in devops?

Good day guys, I've been a developer for quite sometime. I'm interested in the role of devops and started learning courses online. I have successfully deployed on GKE witn Ingress, Helm, Ngix, Docker.

I'd like to know where or how can I find this role. Just a bit background of myself. I've been doing freelance for the past 2 years of my dev career but I wasn't able to get into a team where this techs and methodology is used. Majority of the project I encounter were for individual web app and hybrid apps.

Thank you in advance.

https://redd.it/g3owqs
@r_devops

Update systemd controlled Docker containers

Hi folks,

I'm migrating infrastructure over to docker and I'd like to start the images with systemd as the IAAS provider we are bound to is notorious for restarting machines without warning.

For the production environment I plan to pin the deployed image tags, but for staging, I need to be able to update them automatically through a CI job.

I'm wondering how I could get the automatic updates in place without committing to the infrastructure repository (at least that would feel strange to me). My ideas vary around using sed to replace the systemd unit files and running ansible again with an overwrite-vars file, but both options seem to be cumbersome. Is there a better way I'm missing?

https://redd.it/g3nd01
@r_devops

Looking for a realtime configuration management solution

Hey,

Curious to know what is your experience with configuration management. Looking for a solution that allows for real-time configuration management and monitoring, specifically one that clearly reflects what's it the effective configuration **currently** running per each service on each node.

Also, I would love to know how well do these solutions work with a hybrid environment that runs both a K8S cluster and bare, non-orchestrated applications and services?

https://redd.it/g2hvq5
@r_devops

Did AWS just kill Docker AND Kubernetes with 1 Announcement?

What do you guys think? AWS appears to just have made K8 without the complexities of K8.
[https://medium.com/devops-dudes/did-aws-just-kill-docker-and-kubernetes-with-1-announcement-d7caf5ee4ced](https://medium.com/devops-dudes/did-aws-just-kill-docker-and-kubernetes-with-1-announcement-d7caf5ee4ced)

https://redd.it/g2odpj
@r_devops

MicroK8s?

I have a dedicated and a few VPS systems to which I'm thinking about converting over to MicroK8s. Does anyone have some experience with it? And is it worth the hassle?

https://redd.it/g3wkci
@r_devops

How can we step our game up?

Hi, we are a small dev teams into a small business but we would like to step our game up. We are using devops obviously and we would like use it the best way we can to step our game. We use branches policies that force 2 senior to do code reviews with pull requests. We have development, staging, production environments. We recently added whitesource builds that look for vulnerabilities in our projects. Is there any tools like sonarqube that we could use in a small enterprise for free? What else could we use to set more validations or just set our game up to exploit devops better?

https://redd.it/g481tt
@r_devops

Docker Images : Reducing Image Size

When getting started with containers, it’s pretty easy to be shocked by the size of the images that we build. We’re going to review a number of techniques to reduce image size, without sacrificing developers’ and ops’ convenience. In this first part, we will talk about multi-stage builds, because that’s where anyone should start if they want to reduce the size of their images. We will also explain the differences between static and dynamic linking, as well as why we should care about that. This will be the occasion to introduce Alpine.

https://www.ardanlabs.com/blog/2020/02/docker-images-part1-reducing-image-size.html

https://redd.it/g46atx
@r_devops

Best bare metal provisioning tools for Dell servers?

I am in search of a tool to do bare metal discovery. Essentially I need an image I can network boot (PXE/Bios or UEFI) which can discover all components in the server (CPU/disks/memory/NIC models/etc), firmware versions, what switch ports everything is cabled up to via LLDP, and report all that back to some sort of inventory tool which has an API. These are all Dell servers with an enterprise iDRAC license but multi vendor support is a plus. Once I have all this data I'd like to build some automation so that I can say build a server with this profile and it will query the API for a server meeting those specs and then go and setup the bios options, RAID, configure the switch ports (in a port channel), add an entry for a PXE install (we currently use cobbler, but open to alternatives), then PXE boot to install the OS/desired config. That would almost certainly be Ansible based.

There are several open source bare metal provisioning tools out there but I've not had any experience with any of them except a little Foreman which seemed to be a huge effort to integrate with our existing puppet infrastructure and other existing tools. Can any of them do what I want, or a slightly modified work flow that can achieve the same end result? LLDP info is critical because we currently rely on recording that manually in our inventory tool when we rack new equipment (before the OS install, at which point it's automatically reported if the OS install supports that) and that manual recorded data is all too often wrong/missing and we have to look for it on the switch to figure out where things are actually cabled.

I've thought about just building a live image that can do what we want, we already have an inventory tool we can use to report to, but that might just be extra work on our part that can be avoided with an existing tool.

https://redd.it/g4bnex
@r_devops

Logging additional metadata

In February, before the lockdown, I presented my **Fast logs** talk at ConFoo. At the end, I had an interesting question I had to cut short because of the timing. This blog post aims to describe the relevant points of the talk, the question, as well as some possible answer.

https://blog.frankel.ch/logging-additional-metadata/

https://redd.it/g4aii5
@r_devops

Coding along to learn how to build APIs

Hey guys, do you have any tips for me for great courses that let you code along to build APIs or applications? I am trying to deepen my understanding in application programming design and am eager to learn more about working with API definitions, swagger and the likes. The problem is that the courses I find are either way too basic and teach school programming ( I do have okeyish programming skills ) or are handing out fully coded examples. I can read that code, but I am hoping for something more in between so I can get more hands on and solve the problems myself instead of reading just an example implementation. Any advice for application coding design is also appreciated.

Thanks in advance! Take care during the quarantine.

https://redd.it/g46uy1
@r_devops

What is the value of logging

Hi everyone,

This is probably a very amateur question, however, I’m just starting my own devOps journey and I’m curious as to why I hear about ‘logging’ so much and the benefits of it. Perhaps I’ve missed a post on this forum somewhere but I’d like to understand from some real world examples as to what value it adds to a business. Has it saved an organisation $X due to Y time it saves to find a problem-do these problems happen that often amongst networks, software etc.,? There’s also a tonne of logging providers out there that you pay to achieve a centralised logging system so there must be some figures out there that have shown benefits outweigh the costs.

Would love some help if possible, thanks all for any advice.

https://redd.it/g466bq
@r_devops

How to start ?

I am facing uncertainty in my current job. I have worked on different technologies but dont have complete understanding. Here is the list of topics which I want to learn/cover but I am not sure in which order I should start.

1. Terraform
2. GCP(currently working on AWS but want switch on GCP)
3. Python (I have done few system side scripting)
4. Kubernates( I have configured production level cluster in EKS.)
5. Linux ( want to lean more about system calls and other core stuff).

https://redd.it/g46u12
@r_devops

Slack chatbot needed?

I am looking for the chatbot on Slack that helps me team automate the repeating tasks:
• Call CircleCI API to build jobs
• Ask CTO permissions to deploy on production
Thanks for your reading.

https://redd.it/g475ax
@r_devops

Redis diskless replication: What, how, why and the caveats

[deepsource.io/blog/redis-diskless-replication/](https://deepsource.io/blog/redis-diskless-replication/)

This excerpt aims to address the working and potential of diskless replication in production environments.

https://redd.it/g43u44
@r_devops

Learning DevOps on the cloud

Brief context. Stuck overseas for work with very limited resources in terms of a physical computer. Dell laptop running WSL2. Dual boot with Ubuntu had far too many issues. Should have a fair amount of time to work on DevOps skills as well as some more DB and software engineering tasks.

Any thoughts over which platform like Digital Ocean, AWS, Linode would be the most cost effective and allow me to workon things like:

Ansible, Graffana, Terraform, Docker, Git (for ansible I'm assuming I need to spin up various servers/droplets, etc to make any realistic workflow. )

And from the coding side I'd like to work more on reddis, mongo, OpenCV/Tensorflow.

Thanks for any assistance.

https://redd.it/g43i1x
@r_devops