Cloud Native
What are the elements of a cloud-native architecture?
A cloud-native architecture comprises several technologies unique to platforms such as Microsoft Azure or Amazon AWS.
Microservices
Unlike monolithic applications that contain every object and function, cloud-native apps consist of small, single-purpose microservices. Each service runs independently — spinning up, delivering a service, and spinning down in an on-demand lifecycle.
A microservice’s small size and scope lets developers quickly build, test, and deploy code changes. Rather than timing bundles of changes for an application’s next version release, upgrades get released when needed.
Containers
Containerization is a feature of cloud-native frameworks that helps applications run efficiently at scale and independently of the underlying infrastructure and operating system.
Rather than running multiple applications and microservices side-by-side on a server, container images bundle a single microservice, any dependencies, and a runtime such as Docker. A container registry stores these images for use across the enterprise.
Container orchestration software like Kubernetes manages container configuration, provisioning, and operations at the scales cloud environments demand.
Serverless functions
Serverless computing is a cloud-native way of providing back-end services. A cloud infrastructure provider or a third-party service will deploy code to the platform that lets customers access storage and other services on demand.
Serverless functions add more modularity to cloud applications, making them more scalable and resource-efficient.
Service mesh
To fulfill their roles in an application, independent microservices must be able to communicate with each other. A service mesh acts as a communications layer, allowing microservices to discover each other and exchange data without knowing anything about the underlying infrastructure.
Service meshes complement container orchestration systems by providing advanced load-balancing capabilities to improve performance, encrypting data between microservices, and enforcing security policies.
Immutable infrastructure
Traditional enterprise computing requires teams of engineers to configure the servers running monolithic software. Mutable infrastructure, however, does not scale.
On the other hand, an immutable server only exists as long as it is fit for purpose. Servers are virtual machines with no penalties for short lifecycles. Rather than changing the server to accommodate changing requirements, cloud-native code moves to a new server and the old one is destroyed.
Immutable infrastructure is more predictable and easier to automate at cloud scales.
Cloud-native development
Tying these elements together requires a collaborative, agile approach to application development. Focused on the rapid delivery of business needs, cloud developers rely on automation to build, test, and deploy code at a scale and complexity far beyond monolithic software.
Cloud development combines development and operations (DevOps) practices such as infrastructure as code (IAC), declarative application programming interfaces (APIs), and continuous integration and continuous delivery (CI/CD) workflows.
5 Benefits of a cloud-native approach
When companies began migrating to the cloud, they frequently applied their existing development practices, which meant they could not take advantage of this new form of computing. The benefits of cloud-native applications emerge once companies embrace the nature of the cloud.
1. Resiliency
Cloud platforms offer a level of resiliency that on-premises data centers can never match. Their distributed infrastructure means that, should local disruptions occur, cloud services can fail gracefully to infrastructure in other regions. Minimal downtime protects customer workloads and ensures continuous operations.
2. Scalability
Cloud services abstract infrastructure through a pay-as-you-go business model. Customers only pay for the compute and storage resources consumed. Moreover, these services can dynamically respond to changing demand so customers always have sufficient resources — no more, no less.
Scalable cloud-native apps use microservices, containerization, and serverless functions to take full advantage of this capability.
3. Cost Efficiency
The subscription-based pricing model cloud providers use creates significant savings for their customers. First, businesses no longer need to build and maintain infrastructure. They also avoid the over-investments required to meet peak demands from their applications. Finally, companies optimize spending by writing scalable code that only consumes peak resources during peak demand.
4. Security and Compliance
Cloud-native development emphasizes agility, speed, and automation, strengthening an organization’s security and compliance. Development, security, and operations teams confirm and enforce compliance in code rather than through manual processes. At the same time, continuous monitoring and auditing systems streamline compliance efforts.
5. Ecosystem and Integration
Cloud-native applications access mature ecosystems of managed service and back-end solutions. Using APIs and serverless functions, developers can add new features through an on-demand model that is just as scalable and cost-efficient as their apps.
What is the difference between cloud native and SaaS?
Although related, the terms cloud native and software-as-a-service (SaaS) differ.
Cloud native is a software development approach that fully leverages cloud infrastructure capabilities.
SaaS is a cloud-based software delivery model. However, a particular provider may not use cloud-native application development practices.
What are cloud-native examples?
To help people understand the value of cloud-native architectures, the CNCF has published more than one hundred case studies representing organizations across markets, geographies, and scales. Here are a few examples you can find:
PNC automates compliance
Like other banks, PNC operates in a tight regulatory environment, so balancing speed, agility, and compliance is particularly challenging.
Implementing DevOps practices had already reduced development lifecycles by an order of magnitude to 37 days.
However, 30 days went into creating slides and attending compliance meetings.
PNC developed CI/CD processes to automate compliance checks and complete code reviews in near real-time.
Elkjøp adopts microservices
When this Nordic retailer decided to replace its legacy e-commerce and point-of-sales systems, it took a cloud-native approach.
Replacing monolithic third-party apps with an in-house microservices-based software architecture helped reduce hosting costs by 70%, speed troubleshooting, and improve security.
City of Montréal containerizes its applications
Canada’s second-largest city ran over a thousand applications on platforms ranging from Linux servers to Oracle databases to mainframes.
To modernize their information architecture, city developers containerized their applications in Docker and Kubernetes wherever possible.
Doing so provides the vendor-neutral portability city procurement practices demand while significantly improving efficiency.
How to become cloud native?
Going cloud-native is as much about culture as technology, affecting every aspect of software development and operations. Starburst’s modern data lake analytics platform can facilitate this transformation. Here are a few things to consider as you develop your cloud-native transition plan:
Data Migration
Migrating data from on-premises data centers to one or more cloud services is daunting. It requires careful planning to identify what to move, extract and prepare for the transfer, and then test and validate the data at the other end.
Starburst puts every data source at your fingertips through a single point of access. Migration teams can use our solution to evaluate on-premises data, efficiently develop pipelines, and assess data in cloud storage locations worldwide.
Scalable Data Storage
Startups and established enterprises alike can design robust data lake architectures by adopting scalable data storage solutions, such as Amazon S3 or Azure Blob Storage.
Starburst helps data engineers and consumers alike by delivering fast, scalable data lake access through standard SQL tools.
Data Security and Compliance
Implementing policies in code is the only reliable way to enforce security and compliance. Starburst’s security features include role-based and attribute-based access controls enabling fine-grained policies that limit data access to what users and systems are allowed to see.
Data Catalog, Data Governance and Metadata Management
Despite decades of consolidation efforts, modern enterprises cannot eliminate the inherently distributed nature of data storage. A cloud-native data discovery and governance system streamlines the organization, management, and use of these data assets.
Starburst Gravity’s discovery and governance tools unify an organization’s fragmented storage infrastructure. By creating a single access layer for all data sources, your data teams can easily manage metadata, control access regionally, and enforce consistent governance policies.
Data Storage Cost Optimization
Starburst decouples compute from storage by querying data at the source. Domains can decide how best to manage their storage systems without impacting access for the rest of the organization. Starburst’s cloud-native tools let users run complex queries on-demand without requiring excessive investments in compute resources.
Multi-Cloud and Hybrid Cloud Approach
Before starting their journey to the cloud, companies must understand what “cloud” will mean for their infrastructure. Moving all data and applications to a single provider is an option but may be impractical or unwise. Some legacy applications must remain on-premises, resulting in a hybrid cloud architecture. Other companies will spread their infrastructure across multiple services to create a multi-cloud architecture less susceptible to vendor lock-in. Either approach risks creating new silos that make data inaccessible.
Starburst offers connectors to more than 50 enterprise data sources, letting users access data no matter where it lives. Moving to a new vendor becomes less traumatic. Switching connectors happens seamlessly in the background and is completely transparent to data users.
Self-Service Analytics
Starburst Galaxy democratizes data access by enabling a self-service model. Analysts and data scientists can use their existing SQL tools to access any data source. Starburst automatically handles source-by-source variations such as SQL interpretations or data formatting that once required days of pipeline development. At the same time, granular access control rules ensure that more open data policies do not conflict with compliance requirements.