Continuous Delivery Foundation cdCon 2022, 24 Jun 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Keynote: Continuous Delivery 2.0 – Intelligent CD for the Complex Enterprise - Gopal Dommety

Description

For more Continuous Delivery Foundation content, check out our blog: https://cd.foundation/blog/

Keynote: Continuous Delivery 2.0 – Intelligent CD for the Complex Enterprise - Gopal Dommety, CEO, OpsMx

Increasingly complex teams, workflows, and security and compliance concerns are common hurdles for organizations trying to increase the velocity and accuracy of their multi-cloud software releases. Diverse toolchains and the need for high availability increase the demands. Leading organizations are implementing CD solutions that combine role-based access controls, central policy engines, and AI-based intelligence capabilities to rapidly scale their release volume while achieving new standards for release speed and accuracy. In this talk, Gopal Dommety, CEO of OpsMx will discuss the changes driving the way leading enterprises approach software delivery and the emerging solutions that are enabling massive improvements in Continuous Delivery.

A

Good afternoon today,.

A

I will talk to you share my perspective on the three dilemmas that any enterprise that scaling cd deals with the dilemma of choice, which I think one of the previous.

A

Presenters presented the dilemma of control and the delume of intelligence.

A

For some um so.

A

Sorry, my this, this is not working.

A

A

A

All right and sorry about that, so looking at the last five five years, uh innovation has been accelerating. For example, progressive deployments, uh gitops kubernetes has become mainstream and this transformation has been uh fueled by community platforms such as uh you know the ones like spinnaker argo, so big. Thank you to cdf.

A

A

What's going on all right, um so the the okay, the dilemma of choice is.

A

As a scale increases in any any company, the diversity of use cases increase like tools, apps infra delivery models, increase and it's hard, if not impossible, for any vendor or community to be able to support their needs. And so the dilemma of choice is how do we support these delivery mechanisms needed by businesses and.

A

So uh opsimic's approach is to have a an what we call the ops max opencd, and so it allows for the flexibility to have best of cds platforms um so that developers can choose their uh use. Cases and opencd is built on uh spinnaker and argo, and the experience is unified uh in the sense that the workflow choice drives. The underlying cdus.

A

Now the next is the dilemma of control, so dilemma of control deals with control, visibility, governance, so the kind of two extremes on one extreme. You have there's a full isolation, not much needed for control visibility. This is typically in when you start off in small companies, you start and then the other extreme. You have full control, uh no independence, which is either not neither of them is good, so we need a happy uh combination of both.

A

So the question really is: how can we ever have an isolated autonomous team that operates its cd with the central oversight, and this is especially needed for large enterprises, so at ops mx, um we approach this dilemma with with what we call the decentralized control um decentralized control is, uh is implemented with what a construct called devops uh team workspaces. um This is, if all of some of you know, kubernetes name, space, it's essentially kubernetes namespace creates a sort of a virtual distributed system for you.

A

Similarly, a devops stream space creates a virtual distribution devops space for you, so the the devops workspaces enables autonomous teams to have independence without the overhead of operating a separate cd and and then for the central teams. It enables a way to have an oversight and and have the required control.

A

So the third is the dilemma of intelligence, and you know most enterprises have complex processes with approvals, governance, audit steps, um currently the very manually intensive steps and and at scale and speed. When you start sort of scaling your your cd or you need something fast. You have to automate these uh uh steps uh through intelligence.

A

um So um when you look at this sort of people always start with a manual uh doing a lot of the approvals manually, and then we need to get to a fully automated uh and in the middle. There are lots of these uh there's in a crawl, walk, run kind of intelligent steps. So um the main question: if we have to get there, then uh how do we get from where we are, which is mainly manual or pockets of intelligence, that we have custom built? Companies have custom built to a fully automated model.

A

So ops max approach is to have a intelligence, automation, layer, the intelligence, automation, layer essentially is think of it. It starts with a a data collection.

A

It collects data from all your devops tools and then an enterprise can automate initially starting with repetitive tasks of low value, like approvals and so on and so forth, and and then you bring in a more advanced sort of aiml, on top of the data to assess risk, etc, and ours is a modular approach, means that today we have four modules, intelligence modules and it can fit in even with your existing cd there's sort of no repentant place.

A

So in summary, if you want to scale a cd, you typically will have three dilemmas: a dilemma of choice, dilemma of control and dilemma of intelligence, and our approach is to have an open cd, a decentralized control and intelligence automation to essentially help you scale. So thank you.