Transcript
This transcript was autogenerated. To make changes, submit a PR.
Hello everyone.
This is Kaushik sa.
I'm a senior DevOps senior deeply passionate about
cloud native infrastructure, automation and GitHubs practices.
I worked across industries, modernizing deployments using Kubernetes,
Helen and GitLab ca and today I'm here to talk about something I
believe is the future of DevOps.
Which is called gis.
Whether you are just exploring GI tops or looking to scale it across multiple teams,
this talk will help you understand what it is, why it matters, and how real companies
are leveraging it to transform operations.
So let's start.
Why.
We we need to use QS and why GI tops exist in the first place.
Traditional infrastructure management, even with tools like Terraform or Ansible
informs manual triggers, inconsistent environments, and often lacks visibility.
GitHub's flipped that model.
It treats GIT as a source of truth, allowing infrastructure changes
to go through the same review, approval and CICD pipelines as code.
The 2023 state of continuous delivery report shows 52% of organization already
started using GI tops, and adoption is growing because it brings consistency,
automation, and confidence to the table.
In the past projects and in the current project, GIS eliminated that it
worked on de but not on pro scenario.
It unified the workflow.
So I just want to talk about the key fundamentals of GitHubs.
There are like four key pillars that define GitHubs version control.
Git tracks every change to infrastructure, just like app code declared to config.
You declared the research state using AML or h She files continuous
reconciliation agents like Argo cd.
Check if your cluster matches Watson.
Get and fix it if it drifts.
Collaboration.
Now infra can go through pool request and approvals.
You bring dev and ops into one streamlined workflow.
I like to call this infrastructure meets software engineering principles.
And that shift is really powerful.
And I just want to start with the GIS architecture and
you know how the workflows.
Think of it like this a developer makes a change to Kubernetes manifest and gi let's
say, updating a memory limit or rolling out a new app version that triggers a
pr which gets reviewed and approved.
As soon as it is merged, your GitHubs controller picks it
up and updates the cluster.
If something goes wrong, you don't assess a it to the server,
you roll back the GI commit.
It's all versioned auditable and secure.
So we used this model on a multi crusher retail system.
Every environment was in sync with git and even engine fixes, like hot fixes
were dragged just deliverable deployments.
So I just want to share with you like what are the different GitHubs
tools and technologies out there.
There are like primarily three GitHubs tools where you can, explore and
use regarding to your organization.
The first one is Argo cd.
Gives a beautiful UI to see what's deployed versus Watson gi.
Great for teams starting with.
Is more lightweight and integrates nicely with customization and her.
It's great for GitHubs at scale.
Jenkins X combines GIS with CI pipelines and it's great
for integrated dev workflows.
Each tool has its place and the beauty of GI is it's too agnostic.
The core idea says the same.
You can also integrate with the Secrets manager, like Hashi Cup
Walt, or a w Secrets manager for secure config handling also.
So now let's talk about the GitHubs and Kubernetes integration.
So Kubernetes and GitHubs are like peanut and butter.
Like they just work together.
Because Kubernetes is already declared to your just layering
Git workflow on top of it.
You store all your AML files in Git and reconcile them automatically.
You can manage not just apps, but entire infrastructure namespace CRDs, SSEs, all.
Let's talk about the benefits, benefits of GI tops implementation.
So what's the actual impact of GI tops, like 50 to 70% faster deployments, audit
trails built in, like you have built in audit trails, consistent environments,
all over the, like organization, one Command Rollback with Kit Vert.
In Real World project I worked on we reduced post-deployment incidents by 40%
within the first quarter of QAPs adoption.
Why?
Because there was no more guesswork.
Everyone knew exactly what was running and when.
Now I just want to talk about few, use case studies.
Where companies like Spotify adopting GI Tops spot.
Spotify had a real challenge.
It has 200 plus microservices, multi clusters and no
unified deployment process.
They adopted one of the GI Tops tools flux CD then built
Backstage as an internal platform that empowered like developers to
self-serve their own deployments.
The results are like 80% fewer failures, three times deployment speed.
This is what happens when you combine GitHubs with DevX tools.
And I just want to explore another case study.
One of the biggest companies like Capital One story is all about the mood.
From manual to infra to Kubernetes.
And then, embrace Argo.
Now they manage three plus 3000 plus apps, cons, consistently across
environments and get this like the slash deployment times by 90%.
90%. This proves likes isn't just for tech startups, it works at enterprise level.
I just want to talk about some of the challenges you're gonna
face when you're using GI Tops.
It's not always smooth.
Here are like three big areas to take to tackle like a security.
You need to use like a least privilege, and you need to manage securities as well.
Multi use layer report structures templating tools
like Hal, our customization.
Oracle organization, adoption, train.
You need to train teams, integrate with your existing CI and communicate value.
So we started the small app got quick wins and then we spread
across the results internally.
That's why we drove this adoption.
If you.
Try GitHubs and hit any blockers.
I would love to chat after the session.
These challenges are like real, but solvable.
So let's talk about the roadmap, like how we gonna implement the GitHubs.
If you are looking to adopt GitHubs, he is the path I suggest you're gonna
start by assessing your current process.
Pick a pilot app and GitHub tool.
Document your wings, scale it gradually, and measure it like,
the metrics, deployment frequency, MTRs, and incident count.
GitHubs is a mindset shift, not just a tool adoption.
It brings visibility, control, and speed to the infrastructure.
So let's stop.
Talk about the, GitHub's future and how the GitHub's trend will go.
So looking ahead, the future is exciting.
I can see a integration is coming.
Imagine GitHub systems prior predicting deployment failures before they happen.
Like a multi-cloud GI UPS is already here, manage A-W-S-G-C-P and on-prem
together Git UPS is even more beyond intra like, managing dbs A ps s and more.
The GitHubs workflow like working group is pushed toward interop and standardization.
It's becoming the backbone of modern cloud.
Ops I can say.
So those who embrace GIS now won't just improve current processes.
They'll just shape the next generation of infrastructure delivery.
So thank you guys.
Thank you so much for listening.
You can find me LinkedIn in LinkedIn.
If you want, you can cut there and thanks for giving this opportunity.
Thank you.