Why does it seem like GitHub goes down every other week these days?

It is currently having a minor outage: https://twitter.com/isdownapp/status/1654152908399681537
But it is starting to seem really frequent: https://isdown.app/integrations/github?c=1683215760


https://twitter.com/githubstatus


Any ideas why?

https://redd.it/137s18r
@r_devops

Is GitHub less stable this year or just me?

Seems like the service has been having issues almost every month now, which is causing a slowdown/downtime in our org. https://www.githubstatus.com/history

Is it just me? I'm really hoping this isn't an indication of a pattern here...

https://redd.it/137ryrk
@r_devops

I write an article about how to automate a ecs task creation and execution using event rules

Terraform ECS article

https://redd.it/137tcx9
@r_devops

Is it possible to install Rancher to manage an already functioning K8S?

Hey folks...

At my workplace, we're implementing K8S on premise and everything was done manually with Kubeadm, but it's starting to get complex to manage directly in K8S as the environment is growing.

We use RHEL 9 and I'm thinking about using Rancher to help with the administration. Questions:

1 - Can I install Rancher to manage an already functioning K8S?

2 - We use Podman instead of Docker, is that okay with Rancher?

3 - If you have any other suggestions besides Rancher, they're also welcome!

I forgot to mention that we only have a single node cluster... so, rebuilding/altering the cluster is a possibility.

https://redd.it/137wcs2
@r_devops

Secure access to web application deployed in Azure VM using PrivateLink

We have an http/https application running in VM (we could place it in private subnet / VPN ) in azure in our subscription. We would like other clients from different subscriptions / on premise resources to have an access to this application securely, so that them and only them would have an access to this http/https application. I know in a simple way we could use firewalls and IPs whitelist, but I would like to use something like Private Link / Service Endpoint , I've skimmed over the Azure docs on that, but did not find an example for VM deploy apps, most of the example relate to app services or other PAAS resources ...

https://redd.it/138047z
@r_devops

Motivated DevOps Fresher Seeking Entry-Level Position



Hello,

Hope this post meets you well. If you are willing to give a DevOps fresher a chance. I’m here please.
I have hands-on experience working on real world projects and I am continuously improving my skills. It mustn’t be a DevOps position, anything to get my foot in the door.

Thank you 😊

https://redd.it/13827ix
@r_devops

Non-CS STEM grad, bootcamp grad, currently working as SRE, <1YOE, should I get a part time Master's Degree? Or do certifications?

My foundational coding skills aren't as strong as I'd like them to be. I pretty much only have experience working on personal CRUD apps. My current professional work experience as an SRE isn't coding heavy either. If I want to continue down this path and become knowledgeable in the DevOps area, what should I spend my extra time doing? My company can reimburse me for a part time Master's program like Georgia Tech's OMSCS. I also have access to other online Udemy/MOOC courses that can help me get certifications (K8, AWS, Azure, OCP, etc.). I don't plan on quitting, because real work experience is a priority. But I'm kind of lost at what to do next to reach my goal of becoming a real knowledgeable DevOps professional.

Edit: I wanted to add that I'd prefer a structured learning plan. Would like to avoid doing personal projects at the moment if I can. I also feel like I need some more credentials on my resume anyways.

https://redd.it/137w9s3
@r_devops

AWS Load Balancer Controller

Hey Guys!

I was looking forward to ask if someone knew a way to prevent the creation of new load balancers every time an ingress resource is deployed. My solution now is just to trust the teams that they will first create a separate Ingress and load balancer, and then they will attach further ingress resources to the already created load balancer using the Load Balancer Name and Group. This solution kinda works, but it would be nice to use something like an annotation, that prevents the teams to create load balancer (and only create the ingress resource) by changing the name or not using the group every time and ingress is deployed. That way, they will create ingresses, but alb won’t be created.

Thanks!

https://redd.it/13858hk
@r_devops

Will be taking Dynatrace associate certification exam any tips?

I'll be having my certification 2 weeks from now and I'm nervous as hell. Any reccomendation/suggestion to pass this? I've been reviewing a lot. Thanks in advance.

https://redd.it/1384c5s
@r_devops

deployment strategies list

hey everyone. i'm building an app and readying to deploy it. is there a list of deployment strategies for a small dockerized non-k8s app?

i've googled around and most answers are SEO'd blogposts of some devops solutions companies and idk if their answers fit my needs.

like there is canary, blue green and other solutions but are they better for small scale apps? say i just run the app through some jenkins or github actions for testings then build it, push to the cloud and just update the servers with a new image.

https://redd.it/137snqc
@r_devops

How do you maintain/keep track of the service dependancies to quickly identify the cascading impact during an incident?

If a service is degraded (high latency, error rate), how do you folks quickly identify the impact of it? I understand the service mesh provides the topology of ingress/egress calls and gives us the graph of dependancies. Does anyone make use of it?

Also I see some folks make use of the high level architecture diagram, but how do we ensure it is getting updated regularly with new features, services added.

https://redd.it/137rpy8
@r_devops

How do you print the content of a db and its replica with kubernetes?

How do you print the content of a db and its replica with kubernetes? Is there an easy way to do this? I am planning to use bitnami postgres helm chart and I need to be able to quickly test if there's replication and if the load is balanced and all db are used, so how can I do this easily?

https://redd.it/138a7xu
@r_devops

How to use Git with SHA-2

I wrote an article if anyone is interested in using git with a non-broken hash function.

https://medium.com/@v3ai/how-to-use-sha-2-git-repositories-6c2a6ed5d580

Hope it's helpful!

https://redd.it/138bis1
@r_devops

Is it possible to use VirtualBox/Vmware instead of cloud subscription to make VMs and practice devOps(excluding cloud practice ofcourse)

Hello,

I'm not sure if I can risk/afford a cloud subscription immediately, so is it possible to completely replicate a cloud experience with Virtualbox, like making multiple VMs in same subnets and trying to perform tasks on them?

Any other way instead of virtualBox is also appreciated please suggest.

Thank you.

https://redd.it/138et3q
@r_devops

Your interview questions for the company

I would like to know what questions you ask the company you are applying for in your interview.
How do you make sure you know everything about the role, culture, processes, etc.?

E.g. 

What are your HO rules?
How do you plan your work (OKRs?)
Can I choose my laptop?
Do you have regular feedback rounds?

Thanks

https://redd.it/138gajf
@r_devops

Best way to ship code to multiple edge devices on intranet?

We currently use ansible by sending code to one device and then shipping it to other devices in a more slicker manner.

I just wanted to know if anyone from the community has a better solution, would tools like Jenkins make sense or the process any better? Feel free to provide your thoughts

https://redd.it/138e987
@r_devops

Using GPT to Analyze Cloud Security Issues for GCP

As a cloud user, you know how important it is to ensure your cloud environment is secure. With the vast number of cloud security issues that can arise, it's challenging to keep up with the manual analysis and resolution process. That's why I'm excited to share with you my experience using Selefra, a Policy-as-code product that incorporates GPT functionality to help users perform cloud security analysis, cost analysis, and architecture analysis efficiently on Google Cloud Platform (GCP).

Selefra's GPT feature allowed me to analyze my GCP products for security issues in a way that was similar to ChatGPT. By simply executing a command and providing my inquiry, Selefra's GPT functionality provided me with quick analysis and results, making it easier to identify potential security issues and vulnerabilities in my cloud environment.

The installation and configuration of Selefra were straightforward, and I was able to start using the product within minutes. Additionally, Selefra's documentation was clear and easy to follow, making it simple for me to understand how to use the product effectively.

Overall, I highly recommend Selefra to any cloud user looking to enhance their cloud security analysis and resolution process. You can find more information about Selefra on their:

Website: **https://www.selefra.io/**

GitHub: **https://github.com/selefra/selefra**

Twitter: **https://twitter.com/SelefraCorp**

Give Selefra a try and experience a faster, more efficient cloud security analysis process today!

https://redd.it/138fhwl
@r_devops

Providing Temp Elevated Access via CICD Pipeline

We are using both AWS and GCP. Under exceptional circumstance we must provide engineers elevated permissions to production data. We currently do this manually, with an SEM manually performing the changes but sometimes forgetting to revoke later.

I know there are tools for more 'on-prem' solutions available, but before investigating further, I wanted to know what other people in the community are using. We are happy to tailor a CICD pipeline if this is the best solution whilst we move towards an off-the-shelf tool.

What do you use, what do you wish you were using?

https://redd.it/138jl21
@r_devops

Containerless developer environments? devenv.sh

Hey all

We've been developing a tool to make development environments easy without containers, while allowing to deploy the same environment using a container to production.

Containers create a lot of cognitive overhead when it comes to operating a developer environment, but on macOS they also cause a huge performance penalty due to all the layers of virtualization.

I wonder if that resonates well with the rest of developers that it's something that has been missing in the past?

See https://devenv.sh/

https://redd.it/138jg0d
@r_devops

Backstage vs Custom

I am part of a company where it makes sense to have an IDP for our operations because of the sheer growth we are experiencing. So we just got to the big question. Should we extend backstage or make one from scratch?
For backstage we will need to extend it to a high degree because the current plugin ecosystem seems to lack the things we are looking for.

What is your take on this?
Which option did you choose and how is it going for you?
What things would you have done differently if you were to start over?

https://redd.it/138rrvo
@r_devops

An idea about instant infrastructure provisioning tool

So this morning I read some article about a wireguard/vpn provider that can help setup network for specific tech, why not extend the idea to something more broad

so I thinking of something like this:

$ 1nstant project:create project1
using project1

$ 1nstant provider:setup gcp/aws/azure/etc
click link to authenticate to xxx: ...
saved project1/providers-xxx.tf

$ 1nstant provider:use netmaker/tailscale/nebula/zerotier/packetfabric/etc
click link to authenticate to xxx: ...
saved project1/providers-xxx.tf

$ 1nstant network:create net1 bw:1G name:bla gcp:subnet:xxx aws:subnet:xxx
created network on gcp, private subnet: ...
created network on aws, private subnet: ...
both connected via xxx
est. price: xx/month
saved project1/network-net1-xxx-to-xxx-via-xxx.tf

$ 1nstant vm:create vm1 cpu:32 ram:128 region:aws@us-east1 ssd:500 os:ubuntuserver:2204 network:net1
est. price: xxx/month
saved project1/vms-vm1.tf

$ 1nstant vm:create vm2 cpu:32 ram:128 region:gcp@asia-northeast1 ssd:500 os:ubuntuserver:2204 network:net1
price: xxx/month
saved project1/vms-vm2.tf

$ 1nstant vm:ssh vm1
ubuntu@vm1$ ping vm2
64 bytes from vm1.project1 (x.x.x.x): icmp_seq=1 ttl=107 time=43.1 ms
^C
^D

$ 1nstant billing:estimate
xxx/month to gcp link: ...
xxx/month to aws link: ...
xxx/month to xxx link: ...

the rest (creating user, setup DNS/LB, deploying service, etc) can be passed to your own favorite CM tool (terraform/ansible/saltstack/puppet/chef/pulumi/etc)

traceback: https://news.ycombinator.com/item?id=35832047

https://redd.it/138va44
@r_devops