r_devops | Unsorted

Telegram-канал r_devops - Reddit DevOps

270

Reddit DevOps. #devops Thanks @reddit2telegram and @r_channels

Subscribe to a channel

Reddit DevOps

The "Stateful App Storage Trap": We overprovisioned our self-managed Postgres/Kafka volumes for a huge ingestion job, and now we’re stuck paying for empty space.

Hey everyone,

Looking for some realistic engineering perspectives on a storage lifecycle problem that’s turning into a quiet standoff between our platform team and finance.

A few months ago, we had to run a large data re-indexing and compaction cycle on our self-managed Postgres and Kafka clusters running on AWS EBS. To avoid any disk-full incidents during ingestion, the on-call team did the safe thing and increased several EBS volumes from around 500GB to 2.5TB.

The ingestion finished, retention/vacuum jobs ran, and now the actual active data footprint is closer to 400GB again.

The problem is we’re now using less than 20% of the allocated storage, while still paying AWS for terabytes of mostly empty block storage.

Our company recently added Kubecost to audit Kubernetes and infra spend, and every Monday it flags these stateful volumes as high-priority waste. Finance sees the reports and asks why we don’t just shrink the volumes back down.

But as everyone here probably knows, expanding EBS is easy. Shrinking it safely is where things get ugly.

To reclaim the space, the team would have to manually scale down replicas, create smaller volumes, run rsync or restore backups, swap mounts/volume references, and coordinate a maintenance window with possible downtime or replication drift risks. For a critical database tier, the blast radius of touching live storage often feels worse than the savings.

So nothing happens, and the oversized volumes stay there.

How are other teams handling this?

Do you mostly ignore Kubecost/FinOps alerts when it comes to stateful storage because reliability matters more, or has anyone actually found a safer way to shrink/reclaim live block storage?

Is manual migration still the only approach people genuinely trust for this?

https://redd.it/1tqjvno
@r_devops

Читать полностью…

Reddit DevOps

AI tools can make one developer faster. The harder question is whether that speed becomes team throughput.

We've been thinking about AI coding tools wrong at the team level.

Most evaluation starts with individual productivity: does this save a developer time? Fair question. But the company question is different. Does the work show up as something the team can inspect, validate, and build on?

Private AI sessions help the person using them. They don't help the team answer:
- What was the assigned work?
- Did it produce a reviewable PR?
- Did CI pass?
- What did the reviewer actually inspect?
- Can we repeat this workflow?

Without those checkpoints, AI productivity stays invisible to the org.

The useful unit isn't "did AI write code?" It's "can the team see the path from assigned work to validated change?"

We've been running AI runners this way: bounded tasks, isolated execution, PRs, CI evidence, human review. The artifacts are what make it measurable — not the AI's output, but the normal engineering trail.

Example: promrail PR #38 — a failed GitHub Actions run became a reviewable CI fix with commits, CI evidence, and human merge decision. Not magic. Artifacts.

I wrote up the full argument here: https://forkline.dev/blog/ai-engineering-throughput-visible-work/

Disclosure: I work on Forkline, an AI runner platform. But the observation about throughput vs private speed applies regardless of tool.

https://redd.it/1tqdpqu
@r_devops

Читать полностью…

Reddit DevOps

Searching for an older talk from Etsy

A while ago I came across a talk from maybe 2010 or 2011 from two people at Etsy called something like "Deploying to prod 20 times a day at Etsy", and I can no longer find it! It was definitely two guys presenting, and a rather "of the times" part that stood out to me is when one of them says that deploying to production without tests isn't DevOps it's just "r-worded" (don't disagree with the sentiment).

I've been thinking of it recently because I think people need to understand just how long ago companies have really been "doing DevOps".

https://redd.it/1tq7f5o
@r_devops

Читать полностью…

Reddit DevOps

Building and IDP using Backstage
https://redd.it/1tq3ppl
@r_devops

Читать полностью…

Reddit DevOps

If your CI/CD or service-mesh mTLS rides on a public CA, you have 6 weeks

A lot of platform teams set up mTLS or service-to-service identity flows against Let's Encrypt or another public CA because it was the path of least resistance. That option is closing.

LE removed the ClientAuth EKU from the default ACME profile in February. The tlsclient holdout profile sunsets 8 July. DigiCert, Sectigo, GlobalSign moving similarly under the Chrome root program rule effective 15 June.

The fix is moving client-cert flows to a private trust root. cert-manager handles it fine with an internal Issuer, but the decision is which Issuer – and which trust scope – per workload. That's a policy decision, not a tooling one.

Anyone already through this migration – how did you handle the trust scope decision across workloads?

https://redd.it/1tq1bbr
@r_devops

Читать полностью…

Reddit DevOps

I don't think anyone at my company actually knows where all our pii lives

Security asked us a simple question during an audit. Where is all customer pii stored. And the room literally went silent lol. Warehouse. Backups. Old postgres instances. Abandoned s3 buckets. Random notebooks. Exported csvs in who knows how many tools.

The funny part is everyone thought somebody else had visibility into it already. I've been going down the AI DSPM rabbit hole recently because this honestly feels like a losing battle once your company gets big enough.

Especially now that every ai project creates 15 more copies of datasets. How are people dealing with this without going insane?



https://redd.it/1tl8oej
@r_devops

Читать полностью…

Reddit DevOps

How I lost my test database twice because of an unsecured server
https://redd.it/1tl2j4h
@r_devops

Читать полностью…

Reddit DevOps

I am not sure why my website is getting 24 k requests.
https://redd.it/1tklfgr
@r_devops

Читать полностью…

Reddit DevOps

pnpm 11 Might Finally Be a Better Default Than npm
https://blog.prateekjain.dev/you-should-move-to-pnpm-from-npm-now-6e84b6cc7778?sk=fe3913487a067decbc069234e4e2ad1f

https://redd.it/1tkh0i4
@r_devops

Читать полностью…

Reddit DevOps

Built an internal AI assistant four months ago. security just asked what access it has. i have no idea

We shipped an internal assistant about four months ago hooks into slack, confluence, jira, and google drive. users authenticate through SSO, agent acts on their behalf worked fine, people use it daily, no complaints.

security came to me last week asking for a list of what it can access and what scopes we granted i pulled it together and sent it over and then looked at it myself properly for the first time.

confluence is read-write across all spaces google drive is full access jira can create and modify issues across every project i picked those scopes four months ago because they made the integration work i didn't think too hard about it at the time.

security came back with questions i couldn't answer. what happens to the OAuth tokens if we switch vendors is there an offboarding process for the agent who reviews its access what does it actually do during a session beyond what the logs show.

i don't have answers for any of that we have an IAM process for employees and service accounts but nothing that covers this it doesn't fit neatly into either category.

is anyone actually governing LLM agent access formally or is everyone just dealing with it when security asks.

https://redd.it/1tkbh1x
@r_devops

Читать полностью…

Reddit DevOps

Pivoting from support to cloud ops. Looking for a reality check.

I have spent the last few years working in operations and support with a heavy focus on escalations and operational excellence. I am currently finishing up my AWS SAA to pivot into a full cloud ops role. I already hold my CCP and FOCP certifications.

I am looking for guidance on how to break into this space, ideally in a remote capacity. I want to find a position that allows me to build my skills without the extreme, high-stress cycle I am used to in my current operations background. I know that passing a test is a totally different planet compared to what happens in a real production environment, so I am looking for a sanity check from someone actually in the trenches.

If you have been in this space for a while and would be willing to share some perspective here, I would appreciate the insight. Even better, if you are open to a quick 10 minute call to tell me how you got your start and how you navigated finding a role that was sustainable, please shoot me a DM.

I am happy to respect your time and keep it brief. Thanks for any help.

https://redd.it/1tk796n
@r_devops

Читать полностью…

Reddit DevOps

CTO banned prompt changes on Fridays after what happened last month

A coworker pushed a small prompt update just before the shift ended on Friday because she thought it would only change the response formatting.
By Saturday morning our support assistant had started answering refund questions with information pulled from completely unrelated conversations.
Nothing major happened but support got flooded and our poor ops guy Daniel was manually reviewing conversations from a hotel lobby in Denver trying to figure out what changed.

The weird thing is if you looked at dashboards everything looked fine. Response times were good, no errors and costs normal but in the background the output quality was slowly breaking.

Starting to think AI products need a whole new on call mindset.

https://redd.it/1tjy9lb
@r_devops

Читать полностью…

Reddit DevOps

Books about Release Engineering and Management

I'm not sure if this is the right place to ask, but do you know any books or courses that can be helpful in release engineering and management, git tagging and repository branch management, versioning, packaging (including its naming and structuring), and so on?

https://redd.it/1tjna9h
@r_devops

Читать полностью…

Reddit DevOps

Maybe I'm overengineering this, but managing AI workloads in production feels weirdly fragmented right now.

I have:

normal app monitoring
separate GPU metrics
separate prompt/version tracking
separate model evaluation logs
separate cost dashboards
and then random scripts duct-taped between all of them

The actual inference part is becoming easier than the infrastructure around it.

Curious if people are converging on a stack yet or if everyone else also has a pile of semi-connected tooling.

https://redd.it/1tjk09w
@r_devops

Читать полностью…

Reddit DevOps

Name the operational shortcut that proved to be the “most harmless” but came to haunt you later.

Most teams have resorted to the following:

* Let’s skip this documentation for now
* We’ll sort out this script later.
* This is just temporary.

And after six months, it becomes an issue for all of us. Name one such harmless operational shortcut that caused DevOps more headaches than anticipated.

https://redd.it/1tjgbkv
@r_devops

Читать полностью…

Reddit DevOps

Would companies want this? "Collaborative AI management and automation"

Sorry i dont have the luck of getting hired by companies and spend time building things trying to think if a company would need it, so i built this software that listens to github events and triggers Claude SDK. It also surfaces the session in a basic frontend client.
For now its just self-hosteable but I'm working on it. Its also fully configurable, so automated workflows can be configured from yaml using claude plugins, skills, a prompt, etc.

What happens for example from an Issue or a PR a Claude session (or more sessions if configured) is created and everyone with the link could jump in it to work in the background with it if needed. Its not just automated work, it can have a team of humans in the loop. So in a way it enables collaborative AI work and management.

So the question i have is, would companies want this? Or nah?

Also is it niche or sounds good?

https://redd.it/1tqfs0s
@r_devops

Читать полностью…

Reddit DevOps

Cheatsheet on cloud services
https://redd.it/1tq5cg3
@r_devops

Читать полностью…

Reddit DevOps

I built a VS Code extension that generates HIPAA/HITECH/GDPR-aware code documentation for healthcare software — would love feedback

Hey everyone,

I've been working on a VS Code extension called **compliancedoc-healthcare** and just published it to the Marketplace. It's aimed at devs building healthcare software who constantly have to think about compliance (HIPAA, HITECH, HITRUST, CMS, OSHA, NIST, GDPR) but don't want to leave their editor to deal with it.

**What it does:**

You select a block of code and run one of four commands:

\- **Explain** — plain-English breakdown for compliance officers or non-technical reviewers (what data it touches, what risks it flags)

\- **Generate Docs** — auto-inserts a JSDoc block above your function, tagged with compliance rule references like `HIPAA-164.312` or `NIST-800-53`

\- **Suggest Refactoring** — flags PHI/PII handling gaps and gives you a compliant refactored version as a starting point

\- **Audit Report** — full formal audit-style report with regulatory mapping, access control assessment, and sign-off readiness — useful for release gates or evidence packages

**Why I built it:**

Healthcare dev teams often handle compliance documentation as a separate, painful afterthought — spreadsheets, manual JSDoc, back-and-forth with compliance officers who can't read code. I wanted something that lived inside the IDE and produced structured output that was actually useful as audit evidence (not just AI fluff).

**A few notes on what it's NOT:**

\- It's not a substitute for a qualified compliance officer. Output should be reviewed before going into regulatory submissions.

\- The refactored code it suggests is a starting point, not an automatic patch.

\- Free tier is 10 generations/month. Pro is unlimited + custom rules.

**Tech stack:** VS Code extension + async backend using the Anthropic SDK for generation.

Would genuinely love feedback — especially from anyone who works in health tech and deals with this stuff day to day. What's missing? What would make this actually useful on your team?

🔗 [VS Code Marketplace link\](https://marketplace.visualstudio.com/items?itemName=compliance-documenter.compliance-documenter-healthcare)

https://redd.it/1tq6j9p
@r_devops

Читать полностью…

Reddit DevOps

Focus more on Cloud Engineering or dive further into DevOps?

I am currently a DevOps engineer but with the names switching up every couple of years, it is now splitting into platform engineering and SRE and other titles. I recently decided to take a moment to see what I actually like to do so I can specialize properly, and while I liked coding, with the introduction of AI, I really want to use it as a tool and not as an agent that does everything and I review.

I asked around and searched and people told me that Cloud Engineering is more architecture and closer to what I want. Platform engineering (to my knowledge) can either be DevOps with a different name or in simple terms, a mini SWE and DevOps for the internal teams in the org and SRE is what it probably says, Site Reliability Engineer.

The intent of this post is to ask professionals here about the reality of the situation as I haven't been anything other than a DevOps engineer (played with everything I mentioned above but didn't specialize so my knowledge is limited). I like to think more low-level rather than monitor the AI to automate code and prompt it to fix something (prompting is a skill on its own lol).

I think my options is either focus more on the cloud architecture side or try to get closer to platform engineering (unsure what SRE does exactly as every title just gets confusing at this point), but I thought Cloud may be a better fit as it is more architecture and a good start If i ever decide to move to something like cloud security.

Edit: Just in case, If you use AI agents and enjoy using them, so less coding and simply more debugging what it found then I am glad and a little jealous you enjoy what you do, but I simply wasn't happy as I'd like to use it as a helping hand and not an autonomous hand and that's more on me.

https://redd.it/1tq2y2d
@r_devops

Читать полностью…

Reddit DevOps

Everyone in my company is discovering that Agentic Workflow is just CICD workflows

With all the buzz, people are just building the same CICD workflows.

https://redd.it/1tlbyqj
@r_devops

Читать полностью…

Reddit DevOps

Five Clusters. Five Lessons. One Production System.
https://crza.dev/blog/five-clusters-five-lessons/

https://redd.it/1tl7ihk
@r_devops

Читать полностью…

Reddit DevOps

AI agent repeated failed deployments? Local governance is finally enforceable.

I was running some coding agents for unattended workflows, giving them shell access for deployments. The agent rolled back a deployment because of a database migration script failure, then proceeded to retry the exact same failing npm install command 11 times. Just burning tokens on a lost cause.
I was tired of agents ignoring 'never run destructive shell commands on production configs' instructions buried in long prompts. Prompt instructions are just suggestions. We've been trying to treat probabilistic text generators like strict state machines, and 'hope' isn't a deployment strategy. Local AI tooling ecosystem growth finally made deterministic local gating viable at scale for real teams giving agents real permissions.
I built ThumbGate (MIT licensed, open source) specifically to make agent governance verifiable and enforceable locally, with zero cloud reliance or LLM calls for enforcement.
It hooks into the Pre Tool Use execution locally. One thumbs-down creates a Pre-Action Check that physically blocks the mistake before the next attempt, creating permanent deterministic memory.
ThumbGate works with Claude Code, Cursor, Cline, any MCP agent, etc., plus popular IDE plugins and uses vector matching (LanceDB + Thompson sampling) to apply gates intelligently.
Quick Start:
npx thumbgate init
GitHub: https://github.com/IgorGanapolsky/ThumbGate
I wanted shared team lessons, so we built a cloud option with dashboards and fine-tuning exports, but you can get started, hardening your workflows with local enforcement in one command for free. Happy to answer any questions about the mechanism or architecture!

DISCLOSURE: I'm the developer of ThumbGate

https://redd.it/1tkuklp
@r_devops

Читать полностью…

Reddit DevOps

Fly to SF for Free!

I'm doing some DevRel work for a YC startup - Archal(YC S26)
https://m.youtube.com/watch?v=uUv\_93iCQWM&ra=m
I am looking for MAANG Engineers/Unicorn Tier AI Engineers to vet for them
They are even down to fly them out to SF and take thier testimonies
https://www.archal.ai/

https://redd.it/1tkjsm8
@r_devops

Читать полностью…

Reddit DevOps

Agent Observability and what I think

Hey all, I wanted to share a perspective on something I've been thinking about a lot lately.

Traditional APM was built for request-response and AI Agents break that model entirely. Because, most of what's on the market right now is just legacy APM with agent added, and that leaves a gap you really only feel when things go wrong. You can see the agent's intent (what it decided to do) OR the system-level impact (latency, errors, resource usage), but not both in the same trace. Unfortunately, you're flying blind through the exact moments when cost spikes.

I think observability at the agent layer is one of the real problems here. It's not solved yet. But it's defined well enough that you can instrument properly if you start now.

UC Santa Cruz published research on this last year (arxiv:2508.02736). They used eBPF to intercept TLS traffic and correlate what the agent intended to do with what actually happened at the kernel level. Less than 3% overhead. Point being that this is architecturally possible.

About 5% of AI model requests fail in production today (Datadog, April 2026 survey). Sixty percent of those failures are capacity-related, not model errors. So, it's an operational gap. And teams that built agent-layer observability into their setup caught those failures before they cascaded into outages. Teams that didn't had incidents.

If you're building agents, start with OpenTelemetry. If you're buying a platform, ask the hard questions: Does this handle reasoning loops as a first-class thing? Can you see the decision tree as a continuous trace? Does it know the difference between a tool failing and the agent misunderstanding the tool? Can you alert on semantic drift?

Those are the questions that separate something actually built for agents from something that's just adding agent features to traditional APM. Honeycomb published their approach. Langfuse and LangSmith are solid for multi-step debugging. There are about 15 tools competing on this now, most built on OpenTelemetry standards.

My candid assessment is that you're going to be in supervised mode for a while. Your agent still needs human approval, there is no way around it right now. That's not going away in the next two years. If a vendor tells you otherwise, that's a red flag.

Curious if people can share a) what does good agent observability actually look like at your scale? And b) what are you currently missing on the observability side if anything?

https://redd.it/1tket5w
@r_devops

Читать полностью…

Reddit DevOps

A Practical Back End Engineering Roadmap
https://semicolony.dev/roadmap/

https://redd.it/1tk9rk5
@r_devops

Читать полностью…

Reddit DevOps

How do I become valuable in DevOps & Cloud within the next 2 years as a student?

I’m currently at the end of my 2nd year in INFT course and want to build a career in DevOps and Cloud. I’m planning to spend the next 2 years seriously learning and building skills.

Where should I start?

What should I focus on first?

What skills/tools are most important in today’s industry?

What projects should I build to stand out for internships/jobs?

Would appreciate guidance from people already working in DevOps or Cloud engineering.

https://redd.it/1tk6apt
@r_devops

Читать полностью…

Reddit DevOps

My nginx server started failing below 10% CPU. Turned out to be a hidden Linux limit.

My nginx server started failing even though CPU usage was below 10%.

At first I suspected:

CPU bottlenecks
RAM
nginx workers
networking

But the real problem ended up being a hidden Linux file descriptor limit:
LimitNOFILE=1024

Once nginx reached around 1024 open file descriptors, new connections started failing even while the server still looked healthy.

I recorded the whole investigation/debugging process here:
https://www.youtube.com/watch?v=Hkn9\_\_5yYhg

Would honestly be interested to hear if other people here have hit similar hidden Linux/systemd bottlenecks in production.

https://redd.it/1tjsquq
@r_devops

Читать полностью…

Reddit DevOps

How I handle multi-tenancy in my AI memory product.

I've been building an AI memory product for the past year. Permanent memory, knowledge graph, voice. The interesting half wasn't the AI. It was keeping each user's memory isolated when the stack is polyglot.

What I didn't realize at the start: real multi-tenancy isn't a database problem, it's a multi-tenant polyglot database orchestration problem. Different category, different toolset.

Most multi-tenancy content stops at the database. "Use tenant_id, use schemas, use separate databases." That covers 30% of what actually shows up in production. Here's the other 70%.

1. Credentials lifecycle. Every tenant needs their own credentials per data store. Generate, store, rotate, revoke. In a polyglot stack that's four sets of credentials per tenant. Forget to revoke a deleted tenant's Redis ACL once and you'll build this properly forever.

2. GDPR delete across all stores. Customer says "delete user X." Where does X live? Postgres rows, Mongo collections, Redis keys, queues, search indexes, embedding vectors, S3, logs. If your tenant abstraction lives only in the database, deletion becomes archeology.

3. Per-store isolation across the polyglot. Postgres, Mongo, Redis each have their own isolation story, and none agree on what "tenant" means. The routing layer has to exist somewhere. Most teams build it inside the app, then every new service reinvents it.

4. Connection pooling per tenant. Shared pool collapses on a noisy tenant. Per-tenant pools means tuning size, eviction, and lifetime across hundreds. Static sizing wastes connections on quiet tenants and starves loud ones. Dynamic means you're writing a scheduler.

5. Audit per tenant. "Show me everything that happened to Acme's data last week." App-level logs miss direct DB access. DB-level logs miss intent. Build the unified trail with tenant context end-to-end, or accept the question takes a day.

6. Per-tenant Mongo provisioning. Mongo has no clean "create a database for this customer" primitive that handles credentials, indexes, validation, and routing in one call. You script admin commands. Multiply by every Mongo change you ship.

7. Cross-tenant leak prevention. Every new code path is a potential leak. Reporting query, admin tool, ETL pipeline. The only durable defense is making cross-tenant access structurally impossible.

8. The "where's my data" question. Enterprise procurement asks: where, what region, what backups, who has access. If tenants are mixed across shared and dedicated infra, the answer takes a week to assemble. Build the residency view as a primitive, not a report.

What I ended up doing

Built a routing layer that treats tenant as a first-class concept. One identity per tenant, mapped to their Postgres, Mongo, Redis, backups, region. One API to provision, one to delete, one to answer "where is this tenant's data." Took 6 months. The hardest part of the whole AI memory product.

The mental shift that actually mattered: stopping at "our app supports multiple users" and starting to think of the infrastructure as mini-cloud infrastructure per customer. Once you cross that line, the 8 problems above stop being edge cases and start being your roadmap.

What your stack looks like and which of these surprised you most?

tenant dashboard

https://redd.it/1tjln2o
@r_devops

Читать полностью…

Reddit DevOps

Python dev (Django/FastAPI/Docker/K8s) trying to break into DevOps — what should I prioritize, and what are the real problems no one warns you about?

Hey everyone, long-time lurker, first time posting here. Looking for honest advice from people who've actually made this kind of transition.



My current stack:

Python · Django / FastAPI · Docker + Compose · Kubernetes (basics) · Redis / PostgreSQL · Celery / Async · Bash / Linux · RTSP / FFmpeg pipelines / LLMs · YOLO / OpenCV

I've been building backend systems and a full AI-powered camera security system from the ground up — ingestion pipelines, async workers, containerized deployments, the whole thing. So I'm not starting from scratch, but I know my infra/ops knowledge has real gaps.

Now I want to go deeper into the operations side — CI/CD pipelines, infrastructure-as-code, monitoring, cloud, reliability engineering. Basically bridge the gap between "I can Dockerize things" and "I own the entire deployment lifecycle."



What I want to learn next:

CI/CD pipelines end-to-end (GitHub Actions, GitLab CI, Jenkins?)
Terraform or Pulumi for infrastructure-as-code
Proper Kubernetes beyond just "kubectl apply" — RBAC, Helm, Ingress, autoscaling
Cloud fundamentals — AWS or GCP (which is better to start with?)
Observability stack — Prometheus, Grafana, ELK, alerting
GitOps workflows — ArgoCD, FluxCD



Real questions for this community:

1. What order should I learn these in? I've seen conflicting roadmaps. Some say start with cloud, others say master Linux first, others say just go build something and learn as you go.
2. What are the actual painful problems nobody tells you about? Not the beginner stuff — I mean the things that trip up even experienced engineers. The stuff that takes months to unlearn or figure out on your own.
3. Career reality check — I'm coming from a Python/ML background. Will that help me in DevOps roles or will recruiters just not take me seriously because I don't have a traditional sysadmin / infra background?



The real problems I'm already anticipating (want your take on these):

Tool sprawl confusion — Terraform vs Pulumi vs CDK vs Ansible vs Chef — no one agrees and every job posting wants something different. How did you pick one and stick with it?
Cloud costs — I have zero experience budgeting cloud infra and I know this bites everyone at some point. Any war stories?
Debugging distributed failures — logs scattered across 10 services, no clear owner, alerts firing at midnight. How long did it take you to get good at this?
Kubernetes complexity cliff — goes from "simple" to genuinely hard very fast, and tutorials always skip the hard parts. What resource actually helped you get past that wall?
"DevOps is a culture, not a role" — some companies don't even have a DevOps team, it's just dumped on top of dev work with no extra support or title. How common is this really?
Imposter syndrome — coming in as a developer, not a sysadmin, means constantly feeling like you're missing some foundational Linux/networking knowledge everyone else just has. Did this get better?

https://redd.it/1tjdatq
@r_devops

Читать полностью…

Reddit DevOps

Help me develop few intermediate to advanced DevOps projects that simulate real time workflows.

Can someone help me do DevOps projects that'll simulate real world workflows and the issues they'll resolve while working in production. I'm trying to pivot to the DevOps Engineer role from a cloud background. I have done some projects like 2 tier 3 tier scalable applications with AWS cloud, using tools like terraform, docker, jenkins. I'll be thankful if anyone can provide much more advanced projects that'll help me land a decent devops engineer role .

https://redd.it/1tjd38z
@r_devops

Читать полностью…
Subscribe to a channel