Tech topics

What is Hybrid Cloud?

What is Hybrid Cloud?

Overview

Hybrid cloud is an IT infrastructure that combines public cloud services and private cloud resources and enables applications to be managed and ported between them. It creates a tough, flexible platform for running a company’s analytical and computing workloads, because applications can run simultaneously on multiple environments within the infrastructure. A hybrid cloud deployment leverages public clouds and private data centers and offers the flexibility to switch at any time.

The Foundation of a Hybrid Analytics Solution

Explore the many benefits of a hybrid analytics solution, and how object storage is helping pave the way to a robust future as organizations discover more and more value in their structured and unstructured data.

Read the position paper

Hybrid Cloud

Why is hybrid cloud valuable?

IT leaders often must determine whether to use public cloud or on-premises infrastructure for their analytical workloads. When deciding between these two capabilities, they will take into account factors like corporate culture, security concerns, and data governance rules.

On-premises object storage and file systems are increasingly being chosen by leaders when they choose to keep their data and applications on-premises. With object store analytics, IT leaders can deliver a transparent data experience for users who typically don’t care where the data is actually stored.

Hybrid cloud advantages

By implementing hybrid cloud, you can increase flexibility, performance, and scalability. It paves the way for digital transformation while maintaining complete control of the data offering the following benefits:
  • Avoid single point of failure scenarios
  • Escape public cloud lock-in
  • Create a strategy for more predictable pricing
  • Place workloads in the right place for price/performance
  • Own the data and leverage the cloud at the same time

 


How does hybrid cloud work?

A hybrid cloud requires a blend of technologies that enable data analytics operations equally on any of the public clouds as well as on-premises. For example, object stores are fundamental for data storage in many hybrid clouds. Cloud-based object stores such as Amazon S3 buckets are functionally similar to some on-premises object storage offered by Pure Storage, Dell EMC ECS, Cloudian, Scality, MinIO, NetApp, H3C, and VAST Data. So the user experience regarding data analytics is the same, regardless of where the data resides.

Likewise, well-architected applications like databases and query engines can be developed to access data either in cloud or in on-premises. While cloud-only databases require all data be loaded to the proprietary cloud, hybrid cloud databases can:

  • Access data that has landed in external tables in any S3-style bucket, allowing the architecture to use common file formats like TXT, CSV, ORC, PARQUET, and AVRO.
  • Use the object store as its main repository, optimizing for data warehouse-style analytics.
  • Seamlessly execute analytics wherever the data resides, either on-cloud or on-premises.

Essentially, hybrid cloud applications can leverage cloud technologies for on-premises deployments, which offers greater flexibility to the enterprise architect.

 


Use cases for hybrid cloud

Here are a few examples of data analytics at work within a hybrid cloud:

High availability and disaster recovery

Hybrid clouds are a perfect way to remove reliance on a single vendor and therefore a single point of failure. In mission critical applications, companies can instantiate a cloud-based solution while also maintaining an on-premises copy for disaster recovery. Since the architecture that combines compute nodes with object store is similar between on-premises copies and cloud production, both environments can be managed similarly.

Develop once and serve data anywhere

There’s value to your developers if they can write one set of services without having to worry about the specifics of any given cloud service-provider. You may uncover subtle differences between providers as your real-life analytical workloads hit the servers. Nuances between complicated JOINs, data loading, short/long queries, or machine learning are bound to surface as you evaluate providers. The “noisy neighbors” problem may also be a factor, as some cloud providers share your resources with other clients and make your SLAs unpredictable. The freedom to move a workload seamlessly to another provider, or even to on-premises, allows you to go anywhere your price and performance decisions take you.

Leveraging previous hardware investments

Your company may have already made a major capital investment in hardware, yet the benefits of cloud architectures, with simplified management tools and scalability, might tempt you to abandon those investments in favor of the cloud. In most cases, you can leverage this commodity hardware for an on-premises private cloud by reconfiguring it, thus creating a hybrid environment with all the advantages described here.

When public cloud isn’t possible

Having a single data infrastructure does not always meet the distinct regulatory and market requirements of each line of business. While cloud technologies may be attractive to your company, there are many reasons you may need to keep at least some data on-premises. For example, regulations may dictate that you keep personally identifiable information (PII) within the country of origin. Configuring an in-country on-premises cloud may be the right solution. On the other hand, sometimes data is so sensitive that only an on-premises solution will do.

 


Getting ready to go hybrid cloud

If a hybrid cloud sounds like a great solution for your enterprise IT environment, there are some steps you can take to determine if a hybrid cloud is right for your organization:

  • Decide which data sets need public cloud or private cloud by the regulations that govern their maintenance.
  • Choose solutions that offer the same data architecture for on-prem private cloud and public cloud, including data loading, analytics, and machine learning.
  • Check compatibility of storage, making sure that whatever object store technology you choose works well with your architecture.
  • Check compatibility with your ETL and BI visualization tools, to ensure data flows throughout your pipelines, and analysts will be able to use their favorite tools.

Footnotes