270
Reddit DevOps. #devops Thanks @reddit2telegram and @r_channels
Mid-level DevOps—final 45-min panel (senior DevOps engineer + IT director) at a compliance/audit firm. HMs and senior DevOps folks, could you please tell me what you'd actually ask in that slot. Thank you!
The company is a cybersecurity/compliance assessment firm—they audit other companies for things like SOC 2, PCI, HIPAA, and FedRAMP—so their own internal infra is held to a high bar. Small, fast-moving team; lots of ownership; on-call and off-hours deploys.
The role covers:
Own/maintain AWS infra with a focus on security, resiliency, observability
IaC (Terraform), CI/CD, containers + Kubernetes
Support dev teams' deployments; automate manual ops via APIs
Support an AI/ML team's infra (model deployment, compute, reproducibility) — some MLOps exposure
Support compliance requirements (SOC 2 / PCI / HIPAA)
Databases (Postgres/MySQL/Redis), Linux, networking
My questions:
1. For a 45-min panel like this, what's the realistic number and depth of technical questions you'd get through?
2. Senior DevOps folks — what's your go-to question that separates someone who actually operates infra from someone who only deploys apps?
3. What do you ask to probe security/compliance instincts specifically (vs generic AWS knowledge)?
4. For the MLOps-adjacent part, what would you expect a mid-level engineer to know vs not know?
5. HMs/directors — in your half of the panel, what are you really evaluating, and what answer makes you a yes vs. a no?
Thanks in advance.
https://redd.it/1ty8czo
@r_devops
How Are You Using AI?
Besides asking it for help debugging issues or providing code templates, is anyone here using AI in a meaningful way at their jobs? I see a lot of posts on AI agents and their capabilities but i havent seen any real world examples of people using AI other than a search engine on steroids.
https://redd.it/1ty50ot
@r_devops
DevSecOps Roadmap
I’m working toward a DevSecOps role and put together this roadmap to guide my learning across cloud, security, automation, and CI/CD. Trying to be intentional about building real-world skills and projects along the way—would love feedback.
---
🧭 DevOps / Cloud / Security Roadmap (Phased Plan)
---
Phase 0 – Foundations
Linux + Bash scripting
Git + GitHub
PowerShell (Windows / AD environment)
Python (automation / scripting)
Logging (Linux syslog / Windows Event Logs)
Git commits (clear messages / branches)
Real-world Git usage (code reviews)
Pull request / branching strategies (Git flow)
Linux process management (ps / top / htop)
Linux permissions & users
Linux systemd
Linux networking tools (netstat / ss / curl / tcpdump)
👉 Milestone Project
---
Phase I – Identity & Access Management + Security
Active Directory
Azure AD (Entra ID)
Okta
Google Workspace
Jira / ServiceNow
IAM fundamentals
MFA + Conditional Access
Zero Trust principles
Security + certs
SC-300 cert
IAM misconfiguration scenarios (privilege escalation)
Practice logging / alerting
👉 Milestone Project
🎓 Certifications
CCNA
AZ-104 / SC-300
AZ-500
Terraform Associate
AWS Cloud Practitioner / DevOps Engineer
CKA
---
Phase II – Databases + Automation + IaC
PostgreSQL (queries, joins, ~150MB datasets)
pgvector (vector DB + text search)
Python (boto3, psycopg2)
Terraform (IaC fundamentals)
Store DB creds securely (no hardcoding)
Secrets management (env vars / Vault intro)
Deeper Python (clean code / advanced scripts)
Build small app (Flask / FastAPI)
Cost awareness (AWS cost elimination)
Use tags in Terraform
👉 Milestone Project
---
Phase III – Containers & AWS
Docker (Dockerfile / Compose)
Kubernetes (Pods / Deployments / Services)
AWS:
IAM
EC2
S3
VPC
CloudWatch
CI/CD pipeline
Least-privilege IAM roles
CloudWatch for suspicious activity
Networking Fundamentals:
DNS
HTTP / HTTPS
TLS
Load balancers (ALB / NLB)
NAT
Routing
Subnets
How traffic flows in Kubernetes
👉 Milestone Project
---
Phase IV – Automation & Configuration
Ansible (playbooks / roles)
Terraform + Ansible integration
Configuration drift detection
Immutable infrastructure concepts
👉 Milestone Project
---
Phase V – CI/CD Pipelines + DevSecOps
Jenkins / GitHub Actions
CI/CD pipelines (build → test → deploy)
Trivy (container scanning)
Snyk / Checkov / tfsec (IaC scanning)
HashiCorp Vault (secrets)
OPA / Kyverno (policy as code)
Azure Security (Defender / Key Vault)
AWS pipelines
LLM security (prompt injection / PII protection)
Pipeline Security:
Fail pipelines on vulnerabilities
Block deploys if insecure
Generate security reports automatically
Observability:
Prometheus + Grafana
Logs: ELK stack / Loki
Alerting & IR:
Alerting basics
Incident response basics
Runbooks (incident scenario → response steps)
👉 Milestone Project
---
Phase VI – Integration + Job Prep
3–5 portfolio projects
Practice Jira-style documentation
Combine everything:
Terraform (AWS + Azure)
Docker + Kubernetes
CI/CD pipelines
IAM
Security scanning
👉 Milestone Project
---
⏱️ Weekly Structure
Day 1–4: Learning + Labs
Day 5: Build project
Weekend: Documentation + GitHub
---
https://redd.it/1txvbbk
@r_devops
What should I learn in order to succeed as a entry level devops engineer?
My official title is cloud engineer, and my salary is higher than market average so I really want to not let go of this opportunity. I am contracted for 3 months and depending on my performance they will offer a full time opportunity or not.
I have done some Kubernetes setup on bare metal at school, know python but have almost no experience with azure devops, terraform, CI/CD or infrastructure automation (they told me im going to "automate" and script with powershell heavily).
Should I then focus on terraform and basic powershell scripts?
If anyone has better idea or tips please let me know
https://redd.it/1txuixd
@r_devops
Jr Devops Opportunity
Hey all, I have just been offered an incredible opportunity to do Junior DevOps for a company as I met a higher up through networking. The issue is, I only have jr sys admin experience. I'm confident I can learn what I need to as I have been informed I will be allowed to leverage AI tools and I have been learning cloud recently as well. Is this a realistic jump or am I in over my head? I usually pick things up quick as well. I'm good at being curious and asking questions as well as being willing ti spend free time grinding! Please let me know if I'm a crazy person or if this is possible! Thank you all!
https://redd.it/1txqndp
@r_devops
Checking what are the VPN client people use ?
Hey Team i just joined a startup and here they are planning for standardization so we need to add some vpn.
So checking what are the type of VPN client people using in there organisation (500+ users), which will be secure, reliable and cost efficient.
Let me know what are the VPN client used by your organization and what's the strength of company and how's the VPN latency and security part and if you do how you manager sharing vpn clients and singing per user etc.
https://redd.it/1txnyz8
@r_devops
Is it a problem if I'm only learning on-prem Kubernetes and never touch AWS/Azure?
I'm a junior DevOps engineer and I'm a bit worried about the direction I'm learning in, so I wanted to get some outside opinions.
At my job (and in my personal projects) I work almost entirely with **on-prem / self-managed infrastructure**. The stack I'm learning is roughly:
* **K3s** (self-managed Kubernetes on VMs)
* **Cilium** as the CNI (incl. Gateway API)
* **ArgoCD** for GitOps
* **Ansible** for provisioning
* **Terraform**
* **Longhorn** for storage, **CloudNativePG** for Postgres
* **etc.**..
The thing is, I've **never used a public cloud** — no AWS, Azure, or GCP. No EKS/AKS/GKE, no managed databases, no Terraform against a cloud provider. Everything I do is bare VMs and self-hosted components.
My question: **is this a problem?** A few things I'm wondering:
1. Will I be at a disadvantage in the job market by not knowing the big clouds?
2. Are the concepts I'm learning (Kubernetes internals, networking, GitOps, storage, etc.) transferable to cloud-managed setups, or is it a different world?
3. Should I make an effort to learn a cloud on the side, or is deep on-prem experience valuable enough on its own?
I genuinely enjoy the on-prem / "build it yourself" side of things, I just don't want to accidentally box myself in. Any honest perspective from people who've been in the field longer would be really appreciated. Thanks
https://redd.it/1txhq9k
@r_devops
Looking at Cyberhaven for DLP, curious how it’s been for others
We’ve been looking into Cyberhaven recently while researching DLP options, and trying to get a sense of how it performs in real environments. From what I’ve read, it seems to take a different approach compared to traditional DLP, more around tracking how data moves rather than just enforcing static rules. Conceptually that makes sense, especially with how much work now happens across SaaS apps, endpoints, and AI tools.
If you’ve used it, how does it compare to more traditional DLP tools? Does it reduce noise or just shift it somewhere else? And how difficult is it to get meaningful visibility without a lot of tuning? I’d really appreciate any firsthand Cyberhaven reviews or even secondhand experiences.
https://redd.it/1txf6gk
@r_devops
GitHub - protect Actions yml file from devs
Quick background: we are using Azure DevOps, but migrating to GitHub enterprise for both code repos and deployments. In DevOps all files related to the deployment pipeline are located in the same project, but separate repo. This allows me to control who can modify pipeline files and developers are excluded.
I am having issues achieving the same in GitHub with Actions. There is a .github folder in the repo that I would like to protect. I tried using CODEOWNERS with rules and branch policies. It works, but not as clean as in DevOps. I would like to avoid requiring pull requests for any commit, which is so far the only way I was able to achieve what I want.
Please share how you designed this in your setup.
https://redd.it/1twzp5m
@r_devops
I Built a Retro Terminal Game to Make Kubernetes Less Boring
https://redd.it/1twrooa
@r_devops
Elastic Agent + Kafka: best pattern for routing multiple customer topics to separate indices?
Hey guys, hoping someone with more Fleet/Kafka experience can point me in the right direction here!
We have multiple customers sending data to separate Kafka topics and want each customer's data landing in its own Elasticsearch data stream. We're using the Custom Kafka Logs integration.
I've tried two approaches so far:
\- One integration instance per customer — works, but doesn't feel like it scales well in the Fleet UI - and then the question appearts... will I have 100 kafka integrations on several agents?
\- Single integration + ingest pipeline reroute on `logs-kafka_log.generic@custom` — works for routing, but requires manually updating the pipeline every time a new customer/topic is added, which doesn't feel like the right long-term pattern either
What's the production-grade pattern for this kind of multi-tenant setup? Is one integration per customer actually the way to go, or am I missing something obvious?
Bonus question: we have 4 Elastic Agents across 4 Logstash servers — is increasing topic partitions + shared consumer group the right way to scale consumption across all of them?
Running Elastic Agent 9.3.1 on a 3-node KRaft Kafka cluster. Any help appreciated!
Thanks!
https://redd.it/1twouyt
@r_devops
cracked job interview - applied for dev role, got hired for DevOps skills
https://github.com/harsh-vardhhan/EHR-backend
https://redd.it/1twfh3g
@r_devops
Controlling Telemetry explosion at the Edge with OtelCol and OTTL
https://telflo.com/blog/cutting-observability-costs-at-the-edge-with-ottl
https://redd.it/1twa6i8
@r_devops
Any native Harness templates for OpenClaw or Hermes yet?
Not sure if there is a better subreddit for this but, we are trying to set up an automated release pipeline where an AI agent can review our Terraform plan outputs, check them against our internal security policies, and automatically approve staging deployments.
The problem is we need the agent to run natively within our CI/CD context so it can securely read the repository state and secrets without exposing our infrastructure code to an external API wrapper. I know Harness has some AI features built in now, but does anyone know if there are official pipeline templates or integrations specifically for OpenClaw or Hermes?
Right now we are considering just using gitagent as the runtime to execute the loop inside a standard Harness step. It seems like the cleanest fallback because it lets you structure the agent purely as code and handles the OpenTelemetry tracing. But I would much rather use a native Harness template if one exists to avoid maintaining the custom step ourselves(unless its simpler than I think please correct me there too).
This is a new field with a lot of white gaps and not a lot of material online so any expert advice would help tremendously.
https://redd.it/1tvyjc4
@r_devops
Ai with devops advice
I want some advice about using Ai for DevOps engineer, anyone has a specific setup for agents? Tools? Mcps? Any Ai topic related to DevOps
https://redd.it/1tvw3dn
@r_devops
Linux learning method?
How can I learn linux without actually installing in my PC????
Please Help!!
Thank You!
https://redd.it/1ty6w25
@r_devops
Has undetected Terraform drift ever bitten you in production?
Asking because it happened to us a few months back. Someone opened port 22 to 0.0.0.0/0 during a 2am incident, forgot about it, and then three months later a routine apply silently closed it again. Took us half a day to figure out why things were broken.
I've been poking around for something lightweight that just tells you when your live AWS state diverges from your tf state.
Maybe a morning email report that details what changed by who and how to fix it?
Couldn't find anything that wasn't either enterprise-priced or required a full platform migration.
So I reckon I try building it very scrappily lol. Let me know if this would be useful perhaps?
https://redd.it/1ty0lc7
@r_devops
How are you handling AI quality checks in your deployment pipeline?
Wanted to see if anyone at a Seed - Series A startup has found success with AI eval platforms? We’re shipping new/improving existing AI features pretty regularly and our existing workflows are pretty solid except we don’t have much testing or tracing for our AI-generated outputs.
We’re find that even small prompt tweaks or swapping to the newest model can quietly break output quality in ways that don't surface until a user notices. And right now we’ve got nothing automated that catches that before it ships. I've started looking into eval checks as an actual CI step with the hopes we can block merges if outputs fall below some threshold. Obviously a lot of eval platforms out there but haven’t seen many startups our size adopting those tools yet.
Not trying to add a bunch of work to the team but just hoping to get some core testing in place.
https://redd.it/1txuwzg
@r_devops
A PreToolUse hook that deterministically blocks an AI agent's dangerous tool calls locally — poke holes in the threat model
I've been building a local-first enforcement layer for AI coding agents and want this community to break the approach before I trust it further.
Problem: agents (Claude Code, Cursor, Codex, Gemini CLI, etc.) increasingly run with real shell/tool access. [CLAUDE.md](http://CLAUDE.md) / .cursorrules are suggestions the model can ignore, and most "governance" tooling is observability — it tells you about the `rm -rf` or the `.env` read after it happened.
Approach: intercept at the PreToolUse hook, on the local machine, before the tool call executes. The gate decision is deliberately deterministic — literal pattern match → AST match → scoped rule lookup. No LLM call on the enforcement path, so there's nothing a prompt injection can renegotiate. Where semantic matching is needed (a destructive command not on the literal denylist but close to one we've blocked), it uses local CPU-only bge-small embeddings via LanceDB — no external API.
The part I think is actually different from a hand-rolled hook: a block defined once distributes across every connected agent over MCP stdio, instead of living in one tool's config.
Stuff I'd want critique on:
* Bypass surface: an agent that shells out to a wrapper script to dodge pattern matches. How would you defeat the AST layer?
* Over-firing: there's a short-TTL break-glass recovery path, but I worry about gates trapping the operator.
* Whether deterministic-first is the right call vs a model-judged gate.
MIT, Node >=18. Repo: [https://github.com/IgorGanapolsky/ThumbGate](https://github.com/IgorGanapolsky/ThumbGate)
Not selling anything — genuinely want the failure modes I'm not seeing.
https://redd.it/1txs0bf
@r_devops
Are you ready? http2 bomb
Attack by \~40Mb\/s, oom in 30 seconds
If your nginx/envoy/traefik/haproxy wasn't patched yet, but uses http/2, I advise you to do so before going to weekend.
https://thehackernews.com/2026/06/new-http2-bomb-vulnerability-allows.html
https://redd.it/1txnub0
@r_devops
What's your approach to giving a technical interview post-ai ?
I usually do the standard code challenge where the goal is adhoc log parsing & aggregation. Typically want to see that have at least 1 language (any language) they can write automations in + see/hear their approach. Then a system design call.
I think my system design call is fine but in a post-ai world idk what question I can ask that is super easy for the AI to solve & still reasonable for an interview.
Curious how others are handling this? Bigger more complex challenges?
https://redd.it/1txlmwq
@r_devops
Is Azure capacity this constrained or am I doing it wrong?
I'm working with AWS for many years, and currently I'm working in product with suppose to be cloud agnostic.
I started with AWS and now it's time to spin up it into Azure (because many enterprises using azure for some reason).
I started in US EAST region in azure and at beginning I had an issue with Postgres Flexible, raised a support ticket, and in the result they recommended me to move to another region. The overall conversation to say this takes about 1 day.
I've moved to US EAST 2, and after AKS deployment I stuck with vCPU (Standard Dasv7 Family vCPUs) quote (100) and here we go again... They send me the same message template as they do for previous ticket...
\> ...
\> Your ask for quota has been reviewed and backlogged at this time. It will be reviewed again when additional capacity becomes available. We do not have an ETA for when your request can be fulfilled but please be assured that we will continue working on it and update you as soon as we have more details to share and/or process the request.
\> ...
I'm already waiting for more then 1 day, and there is no responses from their support.
Long Story Short: Because I don't want to wait for days, weeks and months to be able to test infrastructure on Azure. If it will be my decision I just stop and forget about this nightmare. Please suggest the regions and instance types with which I will not have issues.
https://redd.it/1txhaum
@r_devops
Case Study: Building a Betting App on Oracle Free Tier
A client wanted to keep infrastructure costs as close to $0 as possible until the product gets real users.
Current Setup
**Production**
* Oracle ARM VM (2 vCPU, 12 GB RAM, 100 GB Storage)
**Database**
* Separate VM (1 vCPU, 8 GB RAM, 50 GB Storage)
* Connected via private IP only
**Development**
* Separate VM (1 vCPU, 4 GB RAM, 50 GB Storage)
**Backups**
* Database dumps twice daily to Oracle Object Storage
# CI/CD
Developer → GitHub → GitHub Actions
* Code quality checks
* Secret scanning
* Docker build
* Trivy image scan
* Push to GitHub Container Registry
* SSH deployment
* Blue/Green deployment
* Smoke tests before traffic switch
# Security
* Secrets stored in GitHub Actions Secrets
* Database not publicly accessible
* SSL handled by Caddy
# Future Scaling
* Increase Oracle VM resources
* Move DB to AWS RDS
* Migrate fully to AWS if required
I still sketch infrastructure on paper before creating diagrams, so I've included my original handwritten notes as well.
https://preview.redd.it/t4xnawzjbf5h1.jpg?width=1215&format=pjpg&auto=webp&s=aee8d672208565604eb853927c962a4203e1c5d7
**Any improvements you'd make here?**
https://redd.it/1txes8a
@r_devops
Thinking about building an AI on-call investigation tool - talk me out of it / tell me what you'd actually want
Hi, Solo dev here. I keep getting annoyed during on-call at how long the *investigation* part takes - correlating the alert with logs and recent code changes before I even know what to fix. I've been tempted to build something that auto-investigates a page and hands me a first-draft RCA to reduce incidents mean time to resolve specially in midnights.
But I also know this space is crowded (Datadog Bits, incident .io, Cleric, Resolve, HolmesGPT, GitHub's Fix-with-Copilot, etc.), so before I waste months I want a reality check from people who actually carry a pager:
\- Is the investigation step genuinely slow for you, or have existing tools already solved it?
\- For those using an AI SRE/incident tool today: is it actually trusted, or do you re-verify everything it says?
\- What's the one thing none of these tools do that you wish they did?
\- If you're on a small team with no dedicated SRE, do any of these even make sense for you, or is it all enterprise-priced?
Happy to hear 'this already exists, don't bother' - that's useful too. Mostly trying to figure out if there's a real gap or if I'm romanticizing a problem that's already handled.
https://redd.it/1twv0fn
@r_devops
I spent a week auditing our addon upgrade debt. Here's what I found.
So last month I actually sat down and tried to figure out how much time we're burning on addon upgrades across our clusters. cert-manager, ArgoCD, Karpenter, Istio, the usual suspects.
Turns out it's about 3 days a month across the team. Which honestly surprised me because no single upgrade feels that bad in the moment. But it adds up because:
1. Renovate opens the version bump PR but that's like 20% of the actual work. The rest is reading through changelogs, figuring out if any CRDs changed, checking what values got renamed, rewriting stuff, and then writing up rollback notes so the on-call isn't screwed if it breaks.
2. We're never actually caught up. By the time we finish one round there's already new versions out for half the stack. So we're always 2-3 versions behind on something.
3. The compound effect sucks. Skip one minor version, no big deal. Skip three and suddenly you're dealing with cascading breaking changes across multiple release boundaries and what should've been a quick merge turns into a full day thing.
4. It's all tribal knowledge. One person knows how to upgrade ArgoCD. Someone else knows cert-manager. If either of them is on PTO when something needs updating it just doesn't get updated.
We've got Renovate, Pluto, and Nova in place. They're great at telling us what's outdated and what APIs are deprecated. But none of them tell us what actually changed in the helm values between versions, or which CRD fields got renamed, or what the rollback path looks like if things go sideways.
I've been looking into whether LLMs could handle the research and migration part of this, basically reading changelogs across version boundaries, detecting value and CRD changes, and generating the actual manifest diffs. Not the deployment side (ArgoCD handles that fine) but the research and rewriting that eats all the time.
Curious how others are dealing with this:
Is the "research phase" of upgrades just pure manual work for everyone?
Anyone tried throwing AI at parsing release notes and mapping changes to their manifests?
If you're running 10+ addons do you just accept the toil or have you found some way to make it less painful?
https://redd.it/1twmhip
@r_devops
How much timestamp drift do you tolerate before it becomes an operational problem?
Spent way more time on this than I probly should have this week
Was trying to reconstruct an incident across a handful of systems. Nothin was experiencing a failure, NTP was running everywhere (or at least it claimed to be), but a few seconds difference between systems was enough to make the sequence of events annoying to piece together.
Kept finding myself second guessing whether event A happened before event B or if I was just looking at clock drift and chasing ghosts.
Not asking from a compliance/audit angle. More from a day to day troubleshooting perspective.
Is this a pretty common problem, or do I need to review my device configs?
https://redd.it/1twnn6d
@r_devops
B.Tech graduate with no practical skills — is DevOps worth pursuing?
Hi everyone,
I'm a 2026 B.Tech CSE graduate with almost no practical skills, and I'm planning to spend the next few months upskilling in Cloud + DevOps.
With all the discussions around AI, layoffs, and the changing tech market, I'm wondering if DevOps is still a good long-term career choice for someone starting from scratch in 2026.
Would you still recommend DevOps to a beginner today? How is the market for junior DevOps/Cloud engineers, and is it a future-proof path?
I'd really appreciate some honest advice from people already working in the field.
Thanks!
https://redd.it/1twfpmy
@r_devops
After the tj-actions supply chain attack I wrote up
the 7 hardening techniques that would have prevented it
The March 2025 tj-actions incident where 23,000 repos had their secrets exposed through one compromised Action stuck with me. Here are the 7 specific things that would have prevented it.
1. Pin Actions to commit SHAs not tags
A tag like u/v4 can be silently moved to malicious code.
A SHA cannot be faked. This one change protected every team that had done it during CVE-2025-30066.
2. Use OIDC instead of stored secrets
Long lived credentials stay valid until manually rotated.
OIDC tokens expire when the job ends. Nothing to steal.
3. Lock down GITHUB_TOKEN permissions
Add permissions: {} at the top of every workflow and grant each job only what it specifically needs.
4. Treat workflow files like production code
Use CODEOWNERS to require security team review on every .github/workflows/ change before it merges.
5. Scan with Zizmor
pip install zizmor && zizmor .github/workflows/ Catches dangerous pull_request_target configs and script injection risks automatically. Free and takes 2 minutes.
6. Mirror critical Actions into your own org
Fork the Actions you depend on so you are not trusting a stranger's account security.
7. Enforce environment gates
Even a compromised workflow needs human approval before reaching production. That pause catches anomalies.
I wrote a full breakdown with before and after YAML examples for each technique here if anyone needs.
Happy to answer questions in the comments.
https://redd.it/1tw4p08
@r_devops
Scale Kubernetes deployments to zero using KEDA
https://mijndertstuij.nl/posts/scale-to-zero-keda-cron-scaler/
https://redd.it/1tvxura
@r_devops
What should I do to be taken seriously in the job market?
I'm an European developer with 6 years of development experience who started coding for fun. One day, I wanted to know how computers do stuff, and, since then, I've been developing my personal projects and just doing stuff because I like to do so.
Naturally, I´ve learnt a lot of 'sysadmin'/'devops(?)' regarding 'skills'. Like, first with a gh action that cloned and restarted my repos in a VPS. Then, I started using Linux, distro-hopping and learning how ilinux/computer work more deeply.
Eventually, I got into OSS and got a home-server. Deployed some stuff in it with docker on debian. Then, I switched to proxmox and started hosting some of my own stuff in it containerized. After that, I got into Nix(OS) and started declaratively defining my systems in my desktop and some of my VMs...
And, for the last year and a half, I've been doing some 'volunteer' developer work at a non-profit which has made me touch high-avaiability/k8s stuff.
I really never did this looking for a job. I really like learning by myself.
But now, I would like to get into the job market, and devops seem like a great path. I mean, I also like development but there's something intrinsically nice about deploying stuff and managing machines.
For the last few weeks, I've tried applying for development jobs but all the replies I get are: either nothing, ignored or a rejection because of my lack of 'real job' experience. I guess my lack of formal education in development also affects these outcomes.
And idk why, I get a feeling that no matter if I had a giant IaC orchestration system with 20 of the most relevant technologies repo in my GH profile, this wouldn't change the outcome.
So, yeah. What could I do about it?
https://redd.it/1tvslen
@r_devops