Unlocking the Cloud Operating Model: Deployment

We are HashiCorp, a company that provides a set of cloud infrastructure automation solutions across all the critical layers to help organizations find a fast path to cloud and accelerate their application delivery process.

And over the years, we had the chance to work with a broad range of companies from born-in-the-cloud startups to some of the world's largest, most complex organizations. We found that the one common theme in their successful digital transformation is to adopt a new operating model, a cloud operating model.

Cloud Adoption is a Secular Trend

Before we dive into the cloud operating model, let’s first take a quick look at the macro trend to level set.

This is not a surprise to anyone. No matter which industry you are coming from, business today is competing on digital experience. Companies keep adapting how they interact with customers, meaning that they need to respond to market needs, launching new features, and making updates faster than their competitors. It adds a lot of pressure on application delivery and forces the change of all the way from application architecture, development process and the underlying infrastructure technologies.

Accelerating Application Delivery

If you take a closer look at the application delivery process. It really comes down to three main areas of changes

First is Volume and distribution of services Companies used to run a few large monolithic applications, now they may hundreds of or thousands of modular microservices supported or developed by independent developer teams.

While we are decomposing applications, these individual services are packaged and deployed as containers. Different from traditional servers, Containers are generally ephemeral and immutable.

And lastly companies have a lot of choices now, they can deploy applications on their private data centers, or deploy to a cloud or multiple clouds.

These changes impacted developers, operators, their choice of tools and the workflows behind. So you need to think holistically about the cloud operating model that brings people, technology and process together

The Effects of Digital Transformation

The basis of a cloud operating model is to start from the generational transition for IT. Companies shifted from largely dedicated servers in a private datacenter to a pool of compute capacity available on demand. enterprises began with one cloud provider, but may expand to more clouds. You may consider leverage the scale and speed benefits provided clouds and deploy applications there to interface with customers or users.

Your core business databases and internal applications may continue to reside on infrastructure in existing data centers. As a result, enterprises end up with a mix of multiple public and private cloud environments.

So many organizations are trying to determine upfront before they start this journey. If we go down this path, we a strategy that allow us to adopt any infrastructure at any time to support future application development, and we want to truly multi-cloud, won’t be locked in.

A dynamic, multi-cloud datacenter. this reality, is the foundation of the cloud operating model.

Reimagining the Stack

When we actually decompose the cloud operating model, it breaks down to 4 layers, and today we are focusing a top layer, the application deployment. Developers interface with this layer. And this shifted from developers going through operators to deploy an application to a dedicated server to developers themselves being able to directly schedule and deploy a great number of services dynamically across the fleet in a more frequent cycle, and this is where the application automation category comes in.

The Cloud Landscape

Navigating this transition made organizations look for the right solution that fits their business requirements.

We may go from vSphere or vCenter for VMs to a range of technologies for different cloud providers, such as OpenShift for private cloud, and EKS from AWS, and GKE from Google. Each solution has its own assumption to applications, and requires different configurations. It is not hard to see that this ultimately results in a fragmented approach. Companies will end up with different tools, different groups of people, different skill sets, and inconsistent and complex workflows to application deployment.

IT Management Complexity: A Few Statistics

To add a bit more validation to this. We can take a look at a few industry trends.

Deployment options are more diverse than ever, and consequently companies typically have multiple environments for their applications. So a mixture of different operating systems, such as Linux and Windows, bare metal, on-premises, cloud, and multi-cloud deployments are becoming more popular with enterprises.

Meanwhile, management complexity remains as a top challenge of container adoption. The diverse deployment environments further complicate the problem because companies want to stitch these disparate environments into a cohesive one, so a standardized application deployment approach becomes imperative. However, the complexity can this good initiative into a deployment nightmare.

Adopting Containers Could be Hard

Let's look closer at the complexity challenge that many organizations have experienced. There are some examples of pain points around tools and people. It could mean taking a long time to deploy and integrate initially. It could be developers taking significant time and effort in rewriting the applications. It could be a steep learning curve especially when you comprehend a complex system with a lot of moving pieces. If companies decide to migrate from on-prem to cloud, expand to another cloud, start cloud bursting, or scale from central to edge deployment, it could mean a long, complex and painful process of moving workloads.

HashiCorp Nomad

This is why HashiCorp introduced Nomad. The Goal of Nomad is to hide the complexity of underlying target environments from operators and developers and focusing on delivering a simple and flexible workload orchestrator that unifies and consolidates all the different workflows to a single one.

Most of our customers started adopting Nomad as a simple container orchestrator when they start on their containerization or microservice journey. And many companies that with brownfield environment also use nomad to orchestrate their existing infrastructure without the need to force containerization or rewrite them. So companies keep a single workflow before, during and after their migration journey.

Orchestrate Any Application

The guiding principle of Nomad is to orchestrate any application. The goal is not limited Nomad into Docker ecosystem, or a particular OS, or only containerized applications. The benefit is to bring core orchestration benefits to all your applications.

Designed to be simple, lightweight, flexible which can be integrated into your existing infrastructure

This allows companies to do incremental application modernization. Companies can avoid a big overhaul or try to reinvent their existing datacenters. They can achieve results fast and also be able to containerize at their own pace.

Simple Container Orchestration

The first use case Nomad users adopted is simple container orchestrator. We summarized the key benefits and values that customers recognized by adopting Nomad.

Nomad is a single binary, less than 20MB, even smaller than a high resolution picture. It can be easily deployed in customer environments. It takes very short proof of concept cycles to move to production fast. For customers that have already adopted other HashiCorp products, like Consul or Vault, it seamlessly integrates with them.

Nomad provides simplest onboarding experience. Customers have told us that it takes 1 hour or even 30 minutes to onboard a developer who is new to the tool.

The last point is critical for companies that don’t plan to hire additional people to build and maintain the platform. We have Nomad customers that are just running 1-person shops to a team of 4 or 5 people—lean teams that can still provide a robust platform to serve hundreds of developers.

Containerized or Non-Containerized Application Orchestration

Many customers started to move non-containerized applications to Nomad after they made good progress on containers, such as Java applications, application binaries, legacy batch workloads

The biggest benefit is customers start to see values/results immediately without rewrites or changing of their applications and existing infrastructure.

Also helped many companies to improve resource utilization by densely scheduling multiple applications to previously underutilized servers

Lastly, even companies that have legacy applications can still achieve zero downtime deployment by enabling the modern deployment strategies that Nomad supports, such as rolling deploys, blue/green, or canary.

A Single Orchestrator for Clouds

When you zoom out and look at the overall ecosystem and end-end application development workflow. Nomad is the orchestrator engine that can be extended to integrate with other core components, such as networking to provide service discovery, and service mesh, secret management, CI/CD pipeline, storage, specialized hardware.

Broad Ecosystem Integration

And Nomad team is continuously building out more integration with top solutions in each category to drive the growth of the ecosystem.

A Unified Workload Orchestrator For Multi-Cloud

Every customer has their unique environments and requirements, you may wonder where Nomad can be deployed? It can be deployed at all major public clouds, private cloud or bare metal servers. The beauty is when you move a workload between on-prem and cloud with Nomad, it is transparent to the developers. So many customers have used Nomad for cloud bursting or edge deployment across a hybrid environment.

Nomad Adoption

When companies consider the Nomad adoption journey. One of the common patterns is that they started with containers, and then moved their legacy workloads to Nomad as a second phase. And as organizations move more workloads with Nomad to production, they started adopting enterprise features to simplify some operations and ensure they run the platform safely. And they scale internally and onboard more groups, they adopted governance and policy as code features to help them achieve better control.

Nomad Packages

In terms of the Nomad commercial offering, on top of our open source, we have enterprise platform and modules to provides the features run your production environment more safely and efficiently and comes with different levels of support from HashiCorp.

Nomad and the Cloud Operating Model

This wraps up this video. We just walked through Nomad as the deployment layer in the HashiCorp cloud operating model.