Help us gather feedback in the annual community driven AWS survey

This is a yearly survey to collect data on the the state of AWS and related services. The data collected will be released publicly as a resource for all. The more people who submit the more meaningful the data will be, which is why I'm posting it here.

This survey is in not supported, sponsored or endorsed by AWS. It is unofficial and is entirely community made.

After using the State of Javascript website many times, Pete wanted something similar for AWS. After asking around at re:invent 2022, he couldn't find anyone who had put something like this together, so he decided to kick it off.

You can help spread the word and fill it out here! Thanks everyone.

https://answersforaws.com

https://redd.it/10kcj20
@r_devops

Best CI Service for Use with Gitea?

Hey all, I'm looking into deploying my own development environment among other tools that I'm currently setting up on Kubernetes (K3s) and was wondering if there have been any new players in the CI space.

Last I checked, Agola and Drone are the most popular. I tried Woodpecker initially, because it's a fork of Drone and I really want something truly FOSS if possible, but immediately had issues with the log output not being shown in my pipelines. I don't know what it's caused by, but the fact that Woodpecker is still far from a v1 release means it'll have a couple of these issues that I'd like to avoid.

Agola looks great, but the docs are still quite slim and don't mention a setup that doesn't rely on volumes. I want the setup to be distributed and want to be certain I can replace SQLite and whatever else Agola uses with Postgres/S3. Does anyone have experience setting that up?

Drone ends up being the default choice. I set it up, and it immediately worked. Still need to up my Containerd fsnotify limit, but other than that it does the job and has the largest ecosystem among the Gitea-supported CI services.

Between Drone and what I currently was able to find, it seems I'll have to either wait for Gitea to complete their own implementation of CI/CD, or Agola/Woodpecker to improve their documentation so I can solve the problems I've had with them.

https://redd.it/10k61p1
@r_devops

Getting put on pip has made my fears turn into a reality and shaken my confidence.

I've been in the industry for 3.5 years. 1.5 years at my current company (consultancy) where I was about to be put onto pip for under performing. Long story short, I've handed in my resignation and got two offers elsewhere.

My issue is that I had fears of this happening when I left my first company of not knowing enough and that fear has become a reality. The two offers I have I'll be one of two engineers. Which means I relied upon and this pip has really shaken me. Which will mean I'll be heav I feel I need to be in a team of engineers so they can support me and I can learn from them.

They also pay around 10% more, which is also worrying me as higher pay means higher expectations. I couldn't even hack my current job so does asking for more money seem silly?

Can anyone provide me with some advice?

https://redd.it/10ketz7
@r_devops

APM / Observability

Can someone please point me to a reseller or Systems Integrator partner that works with different Observability vendors such as Datadog, Dynatrace, New Relic etc. We would like to speak to someone who would be able to compare the pros and cons of these different vendors, and any would have insight into customer feedback on these products. If you used a reseller/partner when choosing your Observability vendor, please share.

Thx.

https://redd.it/10k8wyc
@r_devops

Best practice for building an internal developer platform

Hi guys,

QQ: We have a green-field project where we are planning to build an internal developer platform with backstage, crossplane and argocd. We don't have much experience as everything is new and we are learning as we go. We have some internal discussions on the team, I think the platform idea changes the development processes, therefore such solution needs evangelisation within the organisation and adoption from the teams, which will require a migration to this new approach.

The other part of the team wants to fit the old processes somehow within the platform and I just don't think how it could work. There's some resistance to changes mostly from dev teams and if we tell them here's a new tool, you need to change how you work and now you own your infra might cause a backslash. Suffice to say It would be messy and very complex to fit our current way of doing things to this new concept.

How should we go about it? We are using Azure, should we even have a different subscription, with different architecture? How can we sell this to management and dev teams?

Any hints would be greatly appreciated. Thanks

https://redd.it/10k20bg
@r_devops

Is 80k/year enough for mid level engineering in major cities in Canada?

I been in DevOps for 3 years now, been in IT for 6 years, I make 120k/year remote.

2 years ago I remember finding a lot of remote jobs for 120k+

The last few months I've been getting a lot of LinkedIn recruiters messaging me offering interviews for non-remote roles in Downtown Toronto for 80k/year.

80k a year after tax in Ontario is about 4,500$/month.

Rent in Downtown Toronto is 3k/month for bare minimum places.

I'm not looking for a new job, I'm just wondering what ya'll think of this?

https://redd.it/10kgunc
@r_devops

I recently extended the GitLab director shadow program criteria to include those between jobs (aka layoffs)

I recently extended the engineering director shadow program criteria to include those between jobs (aka layoffs). The free program is a learning and mentoring opportunity to observe someone living the GitLab values of collaboration, results, efficiency, diversity, inclusion & belonging, iteration, and transparency via asynchronous collaboration and being invited to attend most of my meetings for a week.

You can find out more about the program in the handbook: https://about.gitlab.com/handbook/engineering/development/shadow/director-shadow-program.html

https://redd.it/10kh8jn
@r_devops

POV: offered junior devops position, eventually taking over lead responsibility

As the title states, got an offer on hand for a junior devops position. I’ve worked with CI/CD tools and processes but this is an opportunity to get more hands on.

Shop is AWS/Azure and projects are managed in the cloud.

Goal is to eventually delegate all the responsibilities from the lead/senior devops engineer to myself, and I will be POC going forward but not sure how quickly or how slowly that transition will be. I will be supporting a small team of devs that work on project already in the cloud- most of the infrastructure seems to be set up and looks like I may need to help with any future transitions or changes

Outside of this, I plan on getting familiar with the environment and processes. Also getting a cert aws or azure to make it official.

I think I may take the offer but looking for some advice - what to do to best get myself onboard and up to speed. Seems like a ton of information to take in and don’t want to get in over my head …TIA

https://redd.it/10kgchl
@r_devops

Who is responsible for the Kubernetes "infra-part" in your Organisation?

Hi folks,

we are finally at a point that we can enroll K8s-clusters for dev-projects.
Who is managing the infra part like cert-manager, ingress-controller and monitoring-stack in your organisation?
Is there are central Ops/DevOps/Plattform Team? - If yes, how do you manage a multi-Cluster Environment?
Or is it in the hand of the development team?
Are there channels that everbody can participate on the infra stack?

https://redd.it/10k9fd2
@r_devops

Developer guide for best practices (privacy, GDPR)

I need to write a short, straightforward guide for application devs, but wanted to make sure that one didn’t exist already.

Topics would be targeted at the application layer - eg don’t put this in a log, use cloud events for this, encrypt these fields using this tech, don’t ship logs out of cluster, set up this to cleanse logs on the fly (while you chase dev to make the above changes).

Any favorite resources?

https://redd.it/10k8lbb
@r_devops

Should you build or buy your infra? Checklist for calculating the costs

Last week, I made a post in this sub[1\] & a couple of others requesting feedback on my calculation of the total cost of ownership for Apache Kafka when building/managing it yourself vs. buying it.

Based on the feedback (thanks, by the way!), I've updated my blog post[2\] and included the example calculation I posted on Reddit.

Anyway, here's the "checklist" that guided the calculations made in my previous Reddit post, with links at the bottom:

\-

When calculating the TCO, be sure you calculate the cost for each team involved (e.g., if you have separate infrastructure and development teams, consider the TCO for both independently).

#### Up-front costs

software cost & licensing, if applicable
learning & education
implementation & testing (including data migration costs)
documentation & knowledge sharing
customization

#### Ongoing costs

direct infrastructure costs (e.g., hosting & storage)
backup infrastructure costs (e.g., failover & additional AZs)
supporting infrastructure costs (e.g., monitoring & alerting)
maintenance, patches/upgrades, & support
feature additions

#### Team & opportunity costs

hiring to replace the engineers now working with the new software
time spent on infrastructure that could otherwise be spent on core product

\-

[1\]: https://www.reddit.com/r/devops/comments/10g9bk2/feedback\_request\_tco\_calculation\_for\_apache\_kafka (link to the Reddit post in this sub if you're curious)

[2\]: https://blog.mergent.co/building-vs-buying-software-infrastructure-the-true-total-cost-of-ownership (link to the full blog post)

https://redd.it/10k8egr
@r_devops

Struggling at my job

I managed to secure a job somehow by stretching my experience in my resume however I am struggling badly, I have to design an infrastructure in ECS soon. I don't know where to start learning from. So far everyday is a struggle until a day ends. My learning doesn't get me anywhere so I don't feel motivated to do it either. So far I have:

\- Udemy AWS Cloud practitioner course-Terraform Up and Running book I am reading-Linux and Unix System Admin Handbook-Udemy ECS course I started

I try to do this (one of those) after work but I'm not sure if I get anywhere, I either have to catch up quick or will lose my job. What should my focus be on to get some practical knowledge and what should I focus on ?

https://redd.it/10kpbdn
@r_devops

Starting github action in frontend repo when backend repo changes (w/o backend repo knowing the existence of frontend repo)

I was looking at the repository_dispatch webhook event of github actions but this requires the frontend repo to have repository_dispatch from backend repo and backend repo to explicitly notify the frontend repo with the following api call.

curl \
-X POST \
-H "Accept: application/vnd.github+json" \
-H "Authorization: Bearer <YOUR-TOKEN>"\
-H "X-GitHub-Api-Version: 2022-11-28" \
https://api.github.com/repos/OWNER/REPO/dispatches \
-d '{"eventtype":"on-demand-test","clientpayload":{"unit":false,"integration":true}}'

source

Is there a way for my frontend repo to get notified by backend repo without the backend repo explicitly notifying my frontend? ( I could achieve this with scheduled workflow where I poll the backend repo every x hours but I feel like there is a better way on doing this).

https://redd.it/10kmwbk
@r_devops

Trouble Launching Docker on Windows

So I have been trying to get into the DevOps thing over the past two years. I used to have a laptop that ran Docker Desktop perfectly, allowing me to mess with containers, and run Kubernetes using Minikube.

&#x200B;

Now I lost the laptop and bought me a desktop, which has refused to run the docker engine completely. I have tried a number of options, including running clusters with Hyper-V as the driver to no avail. The Desktop runs on legacy BIOS, but I was told this should not be a problem. After a little troubleshooting I realized that Docker Desktop fails to install dockerd.exe on my system, so the Engine cannot start, and neither does the daemon (am I even getting the terms right?), so it looks like I'll have to build from source. I am told though that this is complicated and I may end up with issues even then.

&#x200B;

It has been a seven month journey of troubleshooting with trial and error and I am just about to give up on this. Has any of you ever faced this? Anyone know a workaround.

My computer's specs are:

HP ProDesk 600 G1 SFF (2014)

Intel(R) Core(TM) i5-4570 CPU @ 3.20GHz 3.20 GHz

Installed RAM: 16.0 GB

BIOS mode: Legacy

Virtualization: Enabled

Hyper-V: Running

&#x200B;

I have tried Docker on WSL, KIND, Minikube and a few other steps from the Docker documentation and stack overflow but haven't had any success. When I have to, I typically spin up a cloud instance, which would be expensive for everyday practice. BTW I'd love to sign up for the CKA and CKAD exam later this year, if anyone is wondering.

https://redd.it/10kr6ll
@r_devops

Alternative to Atlassian Jira and Confluence

Dear all,

Can you recommend a viable alternative to Jira and Confluence? Costs are rising everywhere and I was asked to look into cheaper viable alternatives. Any thoughts?

Context:
Engineering org of about 250 people
Current use of Jira is pretty standard, confluence mainly for documentation (private and for emerging concepts which have not made it to the ‘official’ documentation yet) and exchange of information/ thoughts. Users are mainly software architects, enterprise architects, devs, qa, etc.


Thanks

https://redd.it/10ksowi
@r_devops

Reliable managed CI with SLA?

I'm fed up with my Github actions hanging Queued state forever. It is affecting business and our output. Any experience with other providers? Maybe even something with an SLA, so at least we are compensated for their downtime. All I'm looking for is a tool that, right as someone commits, IMMEDIATELY runs CI. Always.

https://redd.it/10kt02n
@r_devops

Tool to use local IDE for remote development

Hi everyone,

I actually plan to pass my development server on server side and dev in remote.
For the first test i use vs code with the remote connection extension and it works well.
But i want to use Jetbrains IDE and Gateway seams not ready... It crash all the time and take the server with im.

So i'am searching for a tools that make a ssh tunnel for use Jetbrains on 'local' folder.
For the moment i test mutagen.io, looks good but not perfectly stable.
Do you know a solution/tool that can suits my case ?


Thanks a lot !
(maybe this post should go en dev not devops, idk sorry)

https://redd.it/10ktydo
@r_devops

From Operations to Dev Ops

Hello all, would like a bit of advice from devops folks in here. I've been in the operation side of things for 14 odd years (mostly incident, change and release management) but I want to move on to DevOps. I have quit my job 6 months ago but I am now ready to start looking for job. Can you give me some tips on what kind of skills I should get into?

&#x200B;

Just a little bit of background - my main skills were troubleshooting and resolving issues in production AWS and on-premise environment. I have AWS Solutions Architect - Associate, some high level knowledge on Kubernetes, and usage of Splunk/Datadog. Zero on programming language. I am finding job hunting a huge challenge for me since I quit as well.

&#x200B;

Hope you can give me advice and thank you very much in advance!

https://redd.it/10kuiw5
@r_devops

How do you add a migration script that should only run once?

I am thinking I just have to create a sh file, put it inside the repo, and then mount the file into the volume by mapping the repo folder into a docker image folder, and then put

ENTRYPOINT /my-script.sh;

inside the docker-compose file and then remove the sh file and the ENTRYPOINT command, but I am wondering if there's a better way.

https://redd.it/10kod31
@r_devops

Advice for a beginner

Hi all.
So I am just learning web development, and can't decide what to go for, frontend, backend or fullstack, as it it defines what to learn from now.

I personally think backend is more interesting for mw, but when I look for it, it's so much more complicated than frontend, in meaning there is too much options. I don't even know which framework/language is better to learn, while learning new one is a long time for me. Maybe I'd go for Ruby with RoR, or think for future and learn something like Rust to use it in desktop development as well, while there are many other options (Iguess JS isn't best for backend?). I am lost.

While in case of frontend it's at least clear what to learn and where to start with. So basically my question is: is it possible to change specialization after some time of work from frontend to backend? Is it easy? Or employers wouldn't count my frontend experience as an experience for that? I hope after 6-12 months I might figure out for myself what is better to learn for backend etc.

If you have any other advice, please do so. You might help me figure out this path.
Share you experience if you want as well :)

https://redd.it/10kx1a2
@r_devops

Is Build Systems DevOps or just Build Systems Engineering?

My take is given building deb/rpm packages & container image artifacts, licenses & key management that's automated across sub components of a large multi C or Java based project being done through repos (which could include build tools like make or maven) executed & orchestrated from pipelines the repos build order and where they become dependencies would suggest it falls under DevOps? Especially with pipelines being core to supporting the complexities of compiling integrate projects and distributing these builds for an efficient CD work environment to the standard Dev, test, prod spaces.

For me it's something I haven't came across before and could not find more than the typical talk of maven, or other build tools. Here I'm asking to focus all the work around having those build tools managed and interoperability in an automated pipeline/cron job of scripts that have encoded business logic for building such a large mutli repository projects.

I know platform development is also being thrown around. To me I'm guess that would be one level before the builds/distribution and that would be writing code (not TF, or infra) for what needs to be compiled to provided developers a foundation to do there work.

Note: this not making considerations of k8s and deployments. This is intended to be a dialog on building large projects and where does build systems (preferably c binary builds) overlap DevOps.

https://redd.it/10kwwe7
@r_devops