On furlough for 3 months, What should I learn.

Hello all, I have been placed on furlough along with a lot of others in my business, from the UK. What are some things that I could learn in that time? In my day-to-day working with AWS e.g. Cloudformation, API Gateway, and so on.


So I decided at this time I'm going to treat Monday to Friday as normal work time,


I've decided on the following so far


\- Docker

\- Getting better at Python
\- Working towards AWS Dev Ops Pro


Got any suggestions for personal development so I can come back to the business stronger,


Thanks in advance.

https://redd.it/ft35tt
@r_devops

Advice on Transition To DevOps

Hi,

I have been working as a QA/Test Manager for nearly twenty years. For the past 4 months, I have worked as a tribe test engineer which has given some good exposure to DevOps and now I am planning to do a career change into a DevOps Engineer.

Please give me some advice on how I should go about learning to become a DevOps Engineer and how long it would take to be job/interview ready.

Many Thanks

https://redd.it/ft34lm
@r_devops

Resource accountability?

How do you handle resource accountability in your organization? I'm working in a small company as the sole DevOps engineer. I'm struggling with handling accountability for SaaS instances we deploy for clients. While the operational cost is low it's often forgotten when the client leaves, as communication doesn't flow in our direction for those events. Some of our business is also very seasonal or engagement based, so we might want to keep the product deployed for that client perpetually. This is also hard to manage.

In short, how do you keep track of the life expectancy of instances requested?

https://redd.it/ftoasx
@r_devops

Terraform and inital app code base

Good Afternoon,


I've built up quite a nice packer + terraform POC to deploy infrastructure on AWS. It's working very nicely and I'm happy with it and the tool. I'm wondering what others do to deploy their initial application code base onto the terraform infrastructure?


Packer seemed like a good idea, however my application requires an EFS mount and connectivity to an RDS therefore I can't see this working quite right.


Any suggestions would be greatly appreciated as usual.

https://redd.it/fto9al
@r_devops

What's the toughest problem that you'RE Still trying to sovle with kubernetes?

Hi All, I was wondering with what hard problems we are still struggling in day to day activities.

In my experience kubernetes & container technology has come pretty far in last couple of years. Installation can be done in few click on major cloud providers & management has become much more easy as well.

https://redd.it/ftjs7r
@r_devops

Setting Open VPN client in debian-9

I want to set up a VPN client similar to [L2TP VPN - Windows 10 configuration][1].

I have installed `openvpn` using

`sudo apt-get install openvpn`

`/etc/openvpn/client.conf` looks like :

```
client
dev tun
port 1194
proto udp

remote 123.456.789.10 1194
nobind
```
and client configurations taken from [OpenVPN Debian VPN][2].

`sudo openvpn --config /etc/openvpn/client.conf`

> Options error: You must define CA file (--ca) or CA path (--capath)
Use --help for more information.

[Here describes][3] how to do this, but client configuration file was not listed there.

**I can not take any key or certificate file from my server**. I just want to create simple client just as its in Windows-10.

[1]: https://support.untangle.com/hc/en-us/articles/231971387-L2TP-VPN-Windows-10-configuration
[2]: https://wiki.debian.org/OpenVPN
[3]: https://superuser.com/questions/1183305/how-to-install-openvpn-client-in-debian

https://superuser.com/questions/1537806/setting-open-vpn-client-in-debian-9

https://redd.it/fsys6m
@r_devops

Continuous Delivery Stats

Just showcasing that the good fight we are all fighting, the firm I work at typically interviews folks on Continuous Delivery maturity. We took 12 months of our findings into book form.

For example, Canary Deployments are hard, 4% of folks we talked to deploy using a canary.

Book:

[https://harness.io/continuous-delivery-insights-2020-ebook/](https://harness.io/continuous-delivery-insights-2020-ebook/)

Always curious about what challenges folks are facing around Continuous Delivery? Think that CD is a compact disc? Me too \[ironically my car has not had a CD slot since 2010..\]. I would love to hear!

\-ravilach

https://redd.it/ftptnh
@r_devops

Capturing dynamic infrastructure in a Infrastructure as code world

I'm looking for someone to help me connect some dots conceptually.

I understand the basics of IOC, but some situations it disconnects for me.

For example, in my application, when I create a user, I want to create a pubsub topic for that user, so that if I need to send a notification to that user, I can publish to their topic and leave the part of the system subscribed to that to deliver the message.

Likewise, if a 'team' is created in my application, I create a topic for that team, which fans out to the members of the team.

All of these things are dynamically created. How would I capture that into IAC? Or is it simply that this describes something that wouldn't be IAC?

Thanks for reading!

https://redd.it/ftqnml
@r_devops

what do you use python for in the devops space?

Looking for ideas to create some new tools for devops using python. Any ideas would be very appreciated. Looking to solve new problems but also interested in what has already been solved using python.

https://redd.it/ftr5mx
@r_devops

Custom Openshift PAAS Alternative with Kubernetes

Hi, first of all, sorry for my poor english, i'm french. I don't know if i got the good approach to do something.

I work for a company which purprose an web application for banking operation.

We host our application and we have multiple separated instance with different version and different parameter of it. It's hosted with the "old way" : A web server + A BDD on baremetal or on virtual machine . No docker or Kubernetes yet.

As we have many instance with different version, different parameter, we had to maintain some "replica" of each instance for the hotline service which permit them to reproduce some bugs whenever a customer call. For now, we want to change this.

We want something which permits the hotline operator to ask with a web form, an instance of the application with their own parameter and **the version to follow** (for example when a tag is updated, their instance is update). After filled all those information, they receive an url which permit them to access to the instance.

To do it, i first dockerize the app, place all the app parameter as environment variable, then i try openshift 3.11 which have a kind of webform that fill a template file according to the filled parameter. This template file have a Deployment, A service and a route (kind of ingress). Openshift can also "follow" a tag and trigger an rollout on deployment when there is a new tag.

I manage to make it work but after facing many problem (prometheus, nfs subpath volume claim bug etc ) due to the old version of kubernetes included in Openshift 3.11, and after seen the painful process to upgrade to Openshift 4, I decided to switch directly to Kubernetes and write custom script and use CI/CD tools.

I installed kubernetes successfully on baremetal.

I made an little flask application, which permit user to fill an yaml template. This template contains Deployment, A service and an Ingress. Then i apply this yaml to the cluster and reply back with the url.

This feature work.

For the "version to follow" feature. I just looked some tutorial about Jenkins (first time doing Jenkins) .

It's seems that the approach is the following :

Scan the git repository, whenever a specific branch is updated, clone it, Build it, Publish it on the registry, then recreate a specific deployment on the Kubernetes cluster.

But, i want this approach

Scan the git repository, whenever a specific branch is updated, clone it, Build it, publish it on the registry, then patch and rollout many deployment with the same label or with the same image version on the Kubernetes cluster.

As i cannot find any documentation with this approach, i want to know if i have the good approach. Can someone help me please ?

PS : We host everything :

GIT server : Gitea

Docker registry : Private hosted one

https://redd.it/ftrc2n
@r_devops

Searching for a Book of Bamboo server/system(?)

My Boyfriend started to work in IT and looks for a book of the bamboo server. we just only found the pdf version (for free, we know, but we want a real book).. Do anyone have a result? maybe a book with this and some other topics?

https://redd.it/ftstw4
@r_devops

How to host your helm repository on GitHub

Hi everyone, over the past couple of weeks I've been working on [a Helm chart](https://github.com/renovatebot/helm-charts) for [Renovate](https://github.com/renovatebot/renovate), an open source project I'm involved in. Previously, I'd have just hosted this chart in the [helm/charts](https://github.com/helm/charts) repository, but now that's deprecated it wasn't clear where the best place to host helm charts were.

I'd seen some discussions about using GitHub Pages to host on [Helm Hub](https://hub.helm.sh/), and some tools like [Chart Testing](https://github.com/helm/chart-testing) or [helm-docs](https://github.com/norwoodj/helm-docs) to ensure high quality charts and documentation. I hadn't seen anything tying it all together in a complete package, so that's exactly what I tried to do.

Check out [How to host your Helm chart repository on GitHub](https://jamiemagee.co.uk/blog/how-to-host-your-helm-chart-repository-on-github/). It walks you through how to setup GitHub Pages and GitHub Actions build and host your helm repository, using best practices.

I'd love to hear how everyone else is doing this as well. And, of course, if you've got any feedback for me too!

https://redd.it/ftrf9y
@r_devops

Kubernetes cluster down on AWS?

Built a cluster on AWS Cloud9 using EKS but the cluster is down even the master node showing not ready..dont know whats going on since i looked at EKS and check on the nodes. All the nodes even say "Not Ready" even master, it was working previously but think have might done something to break it.

​

"AccessDeniedYour worker nodes do not have access to the cluster.

eksctl-eksworkshop-eksctl-nodegro-NodeInstanceRole-1BJZZG66QZ0KN"

​

I troubleshooted using this guide [https://aws.amazon.com/premiumsupport/knowledge-center/eks-worker-nodes-cluster/](https://aws.amazon.com/premiumsupport/knowledge-center/eks-worker-nodes-cluster/) but problem still persists.

https://redd.it/fty2gc
@r_devops

Your Worklife balanace? On call nights and weekends?

Curious about transiting into devOps. I find the tools, tech, and the mindset interesting but concerned about the work/life balance.

Would you guys mind sharing your perspective on this? Is it the typical 40 hours/week? How often are you on-call, every night or weekends?

https://redd.it/ftu35m
@r_devops

GitHub is down again

[https://www.githubstatus.com/](https://www.githubstatus.com/)

​

Getting 500s on their frontend.

https://redd.it/ftueke
@r_devops

Udemy is having a massive sale and my colleague recently published a course on Prometheus

Hi everyone,


Thought I might help him share his new course on Prometheus monitoring. It's 88 lessons and \~10hr of content. It's only $10 USD because Udemy seems to be having a massive sale rn. The thing is that there's only 1 day left in the sale!


The link is his referral link because the course isn't in the Udemy search cache yet but there's no incentive because every course is still $10 because the sale.
If I'm doing this wrong(promoting), please let me know.


Thank you

[LINK](https://www.udemy.com/course/master-devops-monitoring-with-prometheus/?referralCode=37BDCE3060F77F89A1A2)

https://redd.it/fttayu
@r_devops

Struggling where I should find issues in my IaC like CFT, Terraform, Ansible or ARMTemplate

Infrastructure as Code, it doesn’t matter if via CloudFormation, Terraform, Ansible, ARMTemplate or others, are the de facto standard for creating the cloud environments in most enterprises.

If you are using one of these technologies to build your infra today, where would you prefere to have visibility on any potential issues across the many cloud services that you probably use today, such as public-read buckets (S3) that should be private?

1. In the IDE with some IaC issues detection Plugin
2. CI/CD pipeline with some IaC scanning issues detection
3. Only after building the cloud infrastructure
4. Doesn't matter unless the security team flags it

https://redd.it/ftwjp3
@r_devops

Documentation as a Path to DevOps Automation

Interesting post on how documentation can be a helpful step in setting up automation of your DevOps systems: [https://www.transposit.com/blog/2020.04.02-documentation-as-a-path-to-devops-automation/](https://www.transposit.com/blog/2020.04.02-documentation-as-a-path-to-devops-automation/)

https://redd.it/fu0d33
@r_devops

Considering Microsoft platforms: Azure Devops or Devops Server? Most of our data is on prem

To make things easier, we don't currently have an appropriate devops environment and so consider this as a new instance. Most of our data is on prem. Its mostly sql server and a couple of .net applications hosted on our iis server.

Our hope is to build a more modern devops pipeline. However, with most of our data on prem, it makes sense to go with Azure Devops server on first glance. But I also want to consider whether the cloud version might have something we haven't considered.

We have support the hardware requirements on the on-prem and scale as appropriate so that isn't a deal breaker

Given these few background information, what would be your recommendation? The on-prem version or cloud version?

Thanks

https://redd.it/ftr3d3
@r_devops

Anyone use Concourse at their company? What are your thoughts?

The place I just started at uses Concourse for their CI/CD, but I’ve never used it. Seems pretty straightforward to pick up.

One thing I did notice though is that it seems awfully slow, and the system itself is very unreliable

https://redd.it/fsyefh
@r_devops

Kubernetes Custom Resource Metrics

Kubernetes Operators and Custom Resource Definitions (CRDs) manage Custom Resources. We have added support for getting metrics for Custom Resources, such as CPU/memory usage, number of underlying pods created, etc. in our KubePlus API add-on tool ([https://github.com/cloud-ark/kubeplus](https://github.com/cloud-ark/kubeplus)). You can read about it in this blog post:

[https://medium.com/@cloudark/kubernetes-custom-resource-metrics-2a947dd4d954](https://medium.com/@cloudark/kubernetes-custom-resource-metrics-2a947dd4d954)

Looking for feedback on how you might use these metrics in your environments and what additional metrics you might find useful.

https://redd.it/fu6zmn
@r_devops