Help Maintenance Page, What's under the hood?

TLDR;I want to have customers to be mapped to a Maintenance Page during the release while allowing the team to access the web applications for testing.----

Our PHP Laravel application is hosted inside EC2 on AWS behind an Application Load Balancer and runs Nginx as a Reverse Proxy.

I am tying to find the best practices for enabling and disabling Maintenance Mode on our web application. The way I do it now is by having an ansible script to verify a static page that Says "Maintenace time blah blah blah, we will be back soon." on the production servers, then change Nginx config to map requests to the static page, and finally turn off the Backend (PHP Laravel).

This approach works perfectly but we are facing a hard time smoke testing the application, each time we have to turn off the maintenance mode, start testing, and if we find something we turn it back on and deploy a hotfix.

The question is how the process/setup should be under the hood so customers always get maintenance but our team is always able to access the application during release? What are the best practices for maintenance?

https://redd.it/rktx2b
@r_devops

Articles and Essays that show "the industry is shifting to horizontal scaling"

I'm looking for articles, essays, or published work that shows "tech is scaling horizontally".


I would like to provide evidence for any claims before I make them. Best I can find is Google trends for buzzwords like "kubernetes" and "microservices"


Can anyone help me out?

https://redd.it/rl3bq2
@r_devops

Learn programming

Hi everyone,

I want to move a step forward DevOps/SRE world.
Currently I am a senior system administrator, I gained lot of experience with Linux, containers, Kubernetes and cloud

I think it's time to introduce some programming skills, I see a lot of companies are requiring this kind of skill for DevOps/SRE positions.

I'd like to start my programming journey using Go, but it's difficult to find some books/resources for a newbie like me

Do you have any suggestion?

https://redd.it/rkxsos
@r_devops

SWE has eaten ops, DevOps and SRE. <rant>

&#x200B;

I've been working in tech for almost 20 years, I'm 37. I'm self taught and did not go to college.

I started as a minimum wage tech support rep, became sr. tech support, been an I.T. admin, linux sysadmin, SRE and Sr. SRE and now working at one of the biggest tech companies making 250k in total comp.

I'm very fortunate in having gainful employment at a high compensation level (relatively speaking), considering my lack of formal education and meandering path I've made through life (I had terrible drinking problem until age 25, sober 12 years now, and I was an aspiring semi-pro musician throughout my 20s, not always working).

Right now though, I'm extremely discouraged and frankly kind of annoyed.

I've just done several full-length job interviews for SRE roles at companies that I think are cool (well funded startups), and keep getting told the same thing, which is essentially:

"What we really need is an SWE who has the title of SRE".

&#x200B;

I feel like I've shot myself in the foot a few times mentioning that I consider myself to be a Devops/SRE professional and not a SWE in these interviews (after all, that IS the job I am applying for right?) to one person that then raises this as a red flag during the interview debrief or something.

&#x200B;

I do write code. It's just not the main thing I do, and my SWE colleagues are admittedly much better at it than I am, because that is what they spend all their work time doing. But I am much better at pulling everything together (IaC, pipelines, observability) and making sure it's running in a sane , scalable, and repeatable way.

&#x200B;

Perhaps my experience is not representative of the overall market. I feel like there is or was a time not long ago when SRE was fetching more $ than SWE because the breadth of knowledge required. I'm just not finding this to be the case, if you want to be an SRE it doesn't matter if you have years and years of UNIX/Linux, networking, ops etc, background, maybe you are better off just having a CS degree and 1 year of experience, and "knowing" kubernetes.

&#x200B;

/rant

https://redd.it/rl52rb
@r_devops

Help understanding the point of documentation

Will preface by saying I get the point of BRDs and things like relationship diagrams but i dont get the point of things like documenting what fields are in the database since most databases do a great job at having these things properties easily accessible with descriptions, data types, et Al.

Despite this i often have sr managers request to have things documented in areas outside of the system. Just curious when these types of documents come in use since they're always requiring updates and are a considerable lift to maintain and update on top of the dev work.

https://redd.it/rl5xb5
@r_devops

Are there any devops engineers here that are getting 300k+ TC?

Mainly in bay area and NYC.

https://redd.it/rl7mbb
@r_devops

Designing a DevOps tool chain maintenance activity

Hey there,

We have tools like Jenkins, GitLab, Artifactory hosted in different Azure VMs as Docker containers. Every week we have to restart the containers as per a process. In detail, these have to happen for each tool.

1. Check if the local files for the image are up to date with the remote repository
2. If not, pull the latest code.
3. If there is new code, build the image again
4. Docker compose down
5. Restart docker
6. Docker compose up


At the moment, we have done this manually by SSH ing into the server and executing the commands. But I am thinking of designing a automated way to handle this work. I am wondering what is the best way to do this process automatically ? Is it possible to do something like a Jenkins pipeline or do I need to go the Ansible/Rundeck route ?

https://redd.it/rl5o5k
@r_devops

Looking for study partner for AWS SAA

As the title suggests, I am looking for a study partner for AWS SAA certification. Been working in DevOps, AWS cloud since last 6 months, so have go to knowledge of Linux and AWS. Have Udemy account with courses enrolled. Currently based in India.

https://redd.it/rl57sk
@r_devops

audit.log and cloudwatch agent

What timestamp conversion code are you guys using for the auditd log file? Epoch time seems to not be supported, they all come through as one entry with many actual log lines. It seems like the audit log would have been an important one to support out of the box.

https://redd.it/rla4qq
@r_devops

5 super aliases to quicken development workflow

Here are a few aliases I use everyday in my development lifecycle. Hopefully you will all like them too.

&#x200B;

https://medium.com/nerd-for-tech/my-5-super-aliases-6979b948307d?sk=095b152cca95243409796fb62ea23e05

https://redd.it/rl540d
@r_devops

How to become a DevOps without any CS Background?

Hey guys, I have been in the recruiting industry for 3 years and want to get into IT. I have an MBA but no CS background. I love tech and IT generally attracts me. I wanted to know how I can become a DevOps without a CS degree? Is it even possible, and how should I start?
P.S: I am in Canada.

https://redd.it/rkrotz
@r_devops

DevOps Bulletin Newsletter - Issue 30

Hey folks,

DevOps Bulletin - Digest #30 is out, the following topics are covered:

* **🤖 How to do Canary Deployments in Kubernetes**
* **🐛 The history of Log4j vulnerability and how to mitigate it**
* **👀 Build a monitoring stack with Prometheus, Grafana, Loki and Komodor**
* **🤯 How Twitter are processing billions of events in real-time**
* **🗂 Collect logs on AWS with a pre-configured CloudFormation template**

Complete issue: [https://issues.devopsbulletin.com/issues/inside-the-log4j2-vulnerability.html](https://issues.devopsbulletin.com/issues/inside-the-log4j2-vulnerability.html)

Feedback is welcome :)

https://redd.it/rlcz20
@r_devops

BeyondTrust DevOps

Hello, Does anyone know how BeyondTrust Password Safe can be leveraged to be used in the DevOps process?

https://redd.it/rkqw2m
@r_devops

How many micro services does it take to run your App?

Just a quick ask trying to find out how many micro services does your app need to run. That is if you are using micro services. The app am currently supporting uses about 70 different micro services. All I can say, I miss the monolith.

https://redd.it/rkb3ln
@r_devops

What's your greatest DevOps learning in 2021?

For me, this is what I learned,

Jenkins is not the only CI/CD tool
Security is getting more attention and obviously it has to
Kubernetes is winning all over
DevOps is NOT A ROLE, stop calling people DevOps Engineers
The use of AI and automation in your SDLC is so cool
Just using Git and pushing/pulling things is not GitOps

Share yours!

https://redd.it/rlgkwa
@r_devops

LOG4J shell from a devops perspective

Helle dear devops guys,

I want to know as devops guys what are our responsabilities facing log4shell vulnerability? How should we react and what we need to add in our CI/CD pipelines ? In our nexus repo? Etc.
Untill now i just wrote a shell script that defined all our apps that use log4j 2.x and 1.x , and it's up to the architect to decide if we should upgrade or not, but beside that what can i do more as a devops engineer ?

Thank you all,

https://redd.it/rli8q4
@r_devops

Question about blue/green deployments with databases.

I'm unable to fully comprehend best procedure to implement database migration between the two environments. Wouldn't there be data discrepancies between blue and green? Should I be using database replication between these two so the data is always up-to-date? Most blue/green strategies I read deals with applications deployment but lacks on how to deal ⠀with data. Any input is appreciated.

https://redd.it/rl3gx9
@r_devops

tRANSITIONING FROM CLOUD eNG TO sre?

Hey guys, has anyone made the transition from Cloud Eng -> SRE? I've been a cloud eng for about 2 years now and really want to get into SRE, but it seems like a hard field to break into. Anyone have some advice?

Edit: not a clue wtf happened to the title

https://redd.it/rk81i0
@r_devops

Structuring internal documentation or product/project?

We have several projects/products mostly for B2B customers. Internally we use Confluence for documentation all sorts of decisions, processes, architecture, operations guides, research, security, high level, etc..

Our documentation can still be a mess, with many documents covering multiple aspects that makes it difficult to organize.

How do other people structure project documentation that makes it easy for not only your team to find things, but also business people, and devs from other teams so they don't bug you about "where do I find X?"

https://redd.it/rlm2mp
@r_devops

Podman on MacOS doesn't appear to work

I installed podman using Homebrew and I can't seem to start the machine. It just appears to hang

% brew install podman
% podman machine init
% podman machine start
INFO[0000] waiting for clients...
INFO[0000] listening tcp://127.0.0.1:7777
INFO[0000] new connection from to /var/folders/yd/vwdgp72132g5mdffy066r0nc0000gp/T/podman/qemu_podman-machine-default.sock
Waiting for VM ...
2021/12/21 12:43:03 tcpproxy: for incoming conn [::1]:51050, error dialing "192.168.127.2:22": connect tcp 192.168.127.2:22: no route
2021/12/21 12:43:11 tcpproxy: for incoming conn [::1]:51053, error dialing "192.168.127.2:22": connect tcp 192.168.127.2:22: no route

https://redd.it/rlnstd
@r_devops

How to use hetzner load balancer in kubernetes running in hetzner cloud?

I've set up a new kubernetes cluster on hetzner cloud using kubeadm. I have 1 master node and 2 worker nodes, they are up and running. I used Cilium as the CNI plugin and Traefik as ingress controller. The problem is related to load balancer services - how can I use the hetzner load balancer with them? I tried setting load-balancer.hetzner.cloud/name: "load-balancer-1" to the traefik load balancer service, and even specified loadBalancerIP pointing at the created hetzner load balancer.

Nothing seems to work.... Has anyone managed to get the hetzner load balancer working with kubernetes?

https://redd.it/rlowy9
@r_devops