Kubernetes and Cloud Native Associate (KCNA) Certification. Good for someone looking to break into the field?

I am looking to break into the DevOps field in hopes to get a Junior / Beginner level role after working in IT Support and Application/Software Support for 3-4 years.

I stumbled upon the KNCA Kubernetes and Cloud Native Associate Certification from The Linux Foundation and really liked how it’s geared towards beginners. I have been doing a lot of research and searching at job postings and found that Kubernetes is a in-demand skill.

Do you think this certification is worth while? Obviously it’s beginner based and nothing compared to something like the CKA or CKAD certifications… but was wondering if it’s a good talking point on a resume or to get some looks from recruiters/interviewers/. I was thinking of doing the certification and having some projects that use Kubernetes hosted on a GitHub or talking points where I’ve had hands on experience to supplement the learning.

For a Junior/beginner… would you recommend this? The price isn’t an issue for me at the moment. Here is the link for more info:

https://training.linuxfoundation.org/certification/kubernetes-cloud-native-associate/#review_module

https://redd.it/sz28zt
@r_devops

Does anyone know what happened to slack on 2-22-2022 ?

Hey guys, wanted to understand the cause for slack downtime.

https://redd.it/sz9ldm
@r_devops

ProxySQL cluster behavior question regarding max connections (and advantages/experience with clusting)

I'm considering running a proxysql cluster inside kubernetes and be centralized instead of running sidecars in each app's pod so I have more control over the connections.

I want to be able to make sure there's a maximum amount of connections to my db's per application regardless of how many instances of the application (pods) k8s autoscales to. Lets say the db has a quota of 10k connections. If I run proxysql sidecars, whatever max connection number I set scales as the app scales up and down.

If you're running a ProxySQL cluster and have max\_connections for a hostgroup set to 1000, will that be 1000 total or 1000 per node in the cluster?

I can't find docs that describe its behavior well but it kind of seems like the only advantage of clustering is to apply a config to 1 node and have it propagate automatically.

Does anyone have experience with doing something like this?

https://redd.it/szaw8l
@r_devops

Is it tough to put security into DevOps?

I have recently read an article that made me curious about what others think of putting security into DevOps.

Eurostar Article Link: https://huddle.eurostarsoftwaretesting.com/putting-security-into-devops-is-tougher-than-it-looks/

View Poll

https://redd.it/sz9s46
@r_devops

What are the scenarios where Jenkins is more appropriate over a bash script ?

I have been asked to prepare to take lead of more team members and responsibilities would include decisions on tools.

Have skimmed through the Jenkins reference including Pipeline and syntax like stages/steps etc. In my earlier responsibilities around 7-8 years ago, I have used bash scripts to automate build processes. I do not want to use tools just because they are industry standard/everyone is doing it, without actually understanding what problem it solves. The only thing that I could think of Jenkins solving is OS portability of the build process. This in itself isn't a priority for us. We are on MERN.

https://redd.it/szewp3
@r_devops

From this morning's standup

A colleague, describing a meeting with customers: "They were trying to nail us down on a completion date, and we were trying to nail them down on the requirements."

I can imagine the dialogue in that meeting:

"When do you think you will be done?"

"Well, that depends. What exactly do you want us to do?"

https://redd.it/szjryt
@r_devops

Grafana vs New Relic

Hello, I'm a Prometheus + Grafana user and most of my system metrics are centralized at Grafana. I'm looking to hear more opinions about New Relic users to see if it's worth it.

Some doubts I have:

1. What New Relic does better than Grafana + Prometheus?
2. How complex it is to implement a monitoring system in New Relic?
3. What New Relic can do that Grafana + Prometheus can't?

Feel free to add personal opinions about this topic too. Thanks.

https://redd.it/szl59g
@r_devops

Chef Server OSS

I've been looking and trying to understand which versions of Chef Server are open sourced. From their downloads page I sometimes get the Apache-2.0 license information for versions <=12.14.0, but not all the time. I'm having trouble finding definitive information. Is anyone else running the FOSS version of Chef and can point me in the right direction?

Do I have to build it myself instead of downloading from their 'downloads' page if I wanted it for use with a company? Does anyone have ties to the community Slack page and can send me an invite (seems locked down and only by invitation now)?

https://redd.it/szkiiu
@r_devops

How to monitor Prometheus itself

I currently use Prometheus and Alertmanager in a Kubernetes cluster for monitoring. However, if Prometheus itself or Alertmanager go down then this could only be verified.

Any ideas of another system that could be used to monitor these so there is redundancy? I'm looking for something that could integrate with Slack to send messages if Prometheus went down.

https://redd.it/szllrw
@r_devops

Debugging Azure Devops

I'm trying to build my first Azure CI/CD pipeline and to be honest I'm having a nightmare. The pipeline I'm building is just the standard tutorial one to create a storage account in Azure using an ARM template.

I keep getting the error message:

"Unexpected character encountered when parsing template"

I've checked and rechecked my ARM template and my YAML. I use the ARM add on in VS code to check the json and it seems fine. I can also manually deploy the ARM file in Azure and it works no problem.

I'm not really asking for help on the specific error, what I actually need is a good place to go to find out exactly what the error is telling me. I can't find any record of this error in MS docs and the usual places are no help. So any Azure Devops people out there, do you have a go to resource somewhere out there to help you fix problems with pipelines?

https://redd.it/szp010
@r_devops

Looking for an OPEN SOURCE/CROSS-PLATFORM deploy script. What is everyone using now a days?

I'm looking for a scripting language that I can create an installer with. I was wondering what was popular now a days. Something I can run from the linux or windows command line that will unzip packages in the right locations and will read environment variables from a text file and inject those into the appropriate config files, etc.

I'm open to automation with Terraform, Ansible, etc, is powershell my only option or are there better more task specific scripting language I can use?

https://redd.it/szqhs6
@r_devops

What is the best way to manage MySQL users

We have an RDS MySQL database and currently users are being created and deleted manually. We are starting to grow and I would like to automate this process.

&#x200B;

I already have an ansible playbook which adds all the developers SSH keys to our VMs, so ideally I would just build off of that and manage MySQL users with the Ansible mysql plugin

&#x200B;

Even with that plugin I am still unsure how to create and distribute the passwords once a user is created?

https://redd.it/szojut
@r_devops

Understanding Nginx tail latencies

In this article we trace Nginx running on a 80 CPU server as a CDN node in one of the world largest Internet exchange point. We revealed that a ligh-weight monitoring process may cause severe latencies due to the Linux CPU scheduler. During the investigations we had a lot of fun with eBPF and perf.

&#x200B;

https://tempesta-tech.com/blog/nginx-tail-latency

https://redd.it/szsihc
@r_devops

DevOps Case Study for Interview Process

I've been doing some interviewing and been given some pretty bad technical assessments so far. One company I liked so far wanted me to do a Case study instead of the standard technical assessment for their 3rd round.

However, once I got the details on the case study it looks like they want candidates to solve issues they're currently facing as part of the interview process. Essentially I'm expected to research and come up with a solution to one of the issues presented and do a presentation for 1 hour (hopefully a lot of time for questions).

I don't think this is a complete bait and switch for some free consulting as I'm pretty confident there are actual positions to be filled as multiple roles are posted for. However, I feel like it is still free work provided for any candidates they don't move forward with.

Am I looking too much into it?

https://redd.it/szpvi0
@r_devops

Leveraging Terraform state file for creating resources in azure

Im trying to deploy some Azure resources via terraform for a new environment. The original environment was setup via the gui but i was able to import them via terraform import. I can see all the configurations via the state file.

Although i can create resources via terraform, the documentation only lists a few arguments (i.e., name, location, etc). Is there a way that i can take info from the state file and use it to create new resources using the additional arguments? For example, i can create an app service environment v3 with some arguments. However, i imported an app service environment and it has alot more info. How can i add all the additional info (i.e., inbound_network_dependencies and all of the ip addresses) when deploying via terraform? There has to be a way to add all the extra info at deployment.

https://redd.it/szwjn9
@r_devops

flamegraph.com - A tool for uploading, analyzing, and sharing flamegraphs

At Pyroscope (open source continuous profiling) we use flamegraphs extensively to visualize and analyze profiling data. However, one of the worst parts about using flamegraphs for analysis is that they are kind of annoying to share.

We often found ourselves sharing marked-up screenshots which would lose significant portions of information and important functionality for analysis.

As a solution, we created **flamegraph.com** which creates a shareable link to the flamegraph.

We did this using various components from Pyroscope:

backend: pprof conversion APIs, adhoc api, diff calculator endpoint
frontend: FlamegraphRender component

We added 4 main functionalities for the release of flamegraph.com and we hope to improve these and also add more features in the future:

1. Ability to upload single flamegraph
2. Ability to calculate diff between two flamegraphs
3. API for turning a flamegraph into a shareable link (we use this api for our slackbot / vscode extension)
4. Go playground which both runs and profiles code (currently code has to run for at least \~5 seconds)

Our goal is to improve the experience of sharing flamegraphs and make it easier to do as profiling becomes a more common part of developers observability toolkit.

Definitely a lot of possibilities for flamegraph.com so would love to hear about your experiences with profiling / sharing flamegraphs with your teams and how we could help improve that!

https://redd.it/szv5t5
@r_devops

I accidentally got into devops, what now?

So against all odds I passed two rounds of interviews for the largest university in my country which has about 200 people all in all working in IT (beat 70 candidates, woo!) I was selected to be in DevOps specifically spinning up RHEL linux servers across the different campuses, using among others Ansible, Kubernetes and Terraform. Drifting around 2000 servers.

Problem is.. I do not even have an IT degree. I was upfront about this, and very honest during my interview. I have a maybe 10 university courses in IT under my belt, and some experience here and there. Been using Linux for the past ten years as a power user, but I do not feel this qualifies me for the job.. Very familiar with git, command line and general Linux from an user perspective. But I can not for the life of me name all the folders under root nor what they all do.. I do know a little bit about HTTP, protocols, handshakes and so forth, but its been a while.

Perhaps my best trait Is that I love coding (really, really good at Python), good at learning new things and have a master in mathematics. So bashing my head against difficult things is nothing new.

So... What do I do?

Any advice for how to prepare? I start in august. I am prepared to put in the hours to be as prepared as possible.

https://redd.it/t03m5s
@r_devops

how do you keep your monitoring scripts in sync between servers?

I work at a company where we have a lot of Linux servers monitored by nagios/icinga, we often need to make modifications to our monitoring scripts. The issue is that we want to spread these modifications to all the servers we are monitoring.

I have been thinking to create a repository on GitHub for the monitoring scripts. Then setup the git repository on all the servers then create a periodic git pull on all the servers, so that any modification done on the centralized GitHub repository is automatically synchronized to all the servers we are monitoring.

The reason why I wanted a git repository is to keep the history of all the modifications done to each monitoring scripts.

Is that a good idea? Any other good idea to suggest? How are you guys doing if you are in the same situation as me?

https://redd.it/t08g7n
@r_devops

Is there any way to create a free NAT gateway on AWS?

Hey everyone!

I'm currently deploying my first little project on AWS and was a bit sad to see the bill yesterday, which would probably not allow me to run this long term. The app maintains a project website for a chair I work at at uni. I have a Lambda function packaged in a Docker container image on ECR that connects to S3 and RDS and sends requests to different websites.

When reading up on Lambda, I understood it the way that a function can either (1) only connect to the internet (and not access resources within my VPC), (2) only access recources within the VPC but not connect to the outside or (3) do both if you set up a NAT gateway. I believe that I need both, please correct me if I'm wrong. So I set up a NAT gateway which works fine. However, it charges $0.045 per hour, which would amount to around $30 per month, which would be a little much for a project that does not generate any profit (I cannot ask my uni to pay because they insist we should use their 2017 Debian server that hasn't been updated ever since).

I have tried to find a way to decrease the cost of this. This article suggests that you could use a t3.micro to run a NAT instance on it, but it seems like AWS does not want to support this in the future. I assume that it would also be possible to have other Lambda functions create and destroy the gateway every time it is needed, but that sounds very complicated to me and I would like to keep it as simple as possible.

Do you have any advice on what I could do here?

https://redd.it/t07330
@r_devops

Packer experts need your help

I am trying to create a image out of a base image using packer


I am using this

shared_image_gallery {

subscription = "00000000-0000-0000-0000-00000000000"

resource_group = "ResourceGroup"

gallery_name = "GalleryName"

image_name = "ImageName"

image_version = "1.0.0"

}

managed_image_name = "TargetImageName"

managed_image_resource_group_name = "TargetResourceGroup"

&#x200B;

problem is packer is throwing error like i need provide plan info
however this is custome image and doesnt need those details


can any one please help meon this stuck in this issue for long time

https://redd.it/t09ygi
@r_devops

Looking for DevOps/Cloud Engineers in Europe

Hi! I'm not part of the HR in my company, but an employee looking for new team members, as it's really difficult to hire new people in tech.

Do you have some experience in topics related to DevOps, Cloud or Systems and living in Europe? Are you looking for a new experience? Please PM me and I will redirect your application to the required person, which in turn will meet with you through video conference to see if we have found a match.

Thank you for your interest!

https://redd.it/t0d80d
@r_devops