Cloud Data Warehouse

Multi-Cloud Storage

A man in a data center uses a laptop, managing systems for multi-cloud storage

Multi-cloud storage uses two different cloud stacks for storage. The cloud stacks can be comprised of public clouds, private clouds, or a hybrid cloud.

Why is Multi-Cloud Storage Important?

For a business to take maximum advantage of a cloud-based infrastructure, it is prudent to architect the application and data infrastructure to be portable across different clouds. One cloud provider might offer better pricing, better scalability, easier management, or higher performance than another, and you want the flexibility to use the cloud that best aligns with your needs.

It makes sense to store and process data close to where it is created, when possible, to make it productive immediately and to avoid the cost and potential latency of moving data across a network.

The problem with single-sourcing a platform is vendor lock-in. The cost of a cloud migration is high, and if a vendor realizes that you have taken advantage of its unique features, it can start raising prices and profitability at your expense. If you have a dual-sourcing strategy, the vendor is more compelled to keep its pricing competitive.

Not all cloud storage is the same. You may have an application that performs better on one provider’s block storage than another. In this case, it makes sense to give preference to that provider because it benefits customers.

For a global company, flexibility of deployment is important. In some cases, one provider has a data center in a geography that another does not. Keeping your deployment options open is good practice for performance or regulatory reasons.

The Drawback of Multi-Cloud Storage

The main reasons to avoid multi-cloud are centered around greater complexity. With a single supplier, there is only one vendor to manage. Managing multiple clouds is more challenging than using a single provider. Every provider has its own terminology and processes. Fortunately, storage uses a lot of common terminology.

Another drawback is that with multi-cloud, new applications need to be designed and built with portability as a requirement. This can result in some compromise because the applications have to take the lowest common denominator approach to function in different cloud environments.

Multi-Cloud Storage Considerations

Multi-cloud storage is becoming the norm for most medium and large businesses. This section discusses considerations when adopting a multi-cloud strategy.

Storage access from applications should avoid application programming interfaces (APIs) that only exist in a single cloud. This reduces lock-in problems and prevents migration problems down the road. Plenty of common APIs, such as the Open Source Database Benchmark (OSDB) and Server Message Block (SMB), span multiple cloud platforms.

Many cloud data platforms, such as the Actian Data Platform, offer multi-cloud storage under the covers so customers can benefit from consistent storage management and access, irrespective of the storage platform.

Applications can be insulated from differences in deployment by encapsulating them in a container such as Docker or Kubernetes, creating a single executable unit with all the external dependencies in the container. Microservices takes this portability to the next level by wrapping a web services API around the application.

The Benefits of Multi-Cloud Storage

The popularity of multi-cloud storage can be attributed to many of these benefits:

  • Deployment flexibility is a significant reason to adopt multi-cloud storage. With so many cloud storage providers competing for business, it makes sense to use a dual supplier strategy to get the best value and choice of storage for your applications.
  • Keep pace with innovations in the storage space. If applications are architected for portability, migrating to the fastest or lowest-cost storage is more manageable.
  • Data storage flexibility. One vendor may have a data center close to your customers or employees to help mitigate network latency.
  • Data storage has greater resilience if mirrored across two vendors in case one suffers an outage.
  • For disaster recovery scenarios, one site can archive data to others in a different geography to minimize the chance of both being impacted by the same disaster.
  • For regulatory compliance reasons, different providers guarantee the location of your data in a particular country or state.
  • Service level agreements (SLAs) can vary by provider. If your business has applications with extreme performance or availability requirements, one provider may be able to meet that SLA at a lower cost than another.
  • As the business adopts new paradigms such as serverless microservices, workloads and storage can be load-balanced across multiple clouds.

Actian in Multi-Cloud Environments

The Actian Data Platform is designed to operate in multi-cloud environments. The Actian Data Platform handles multi-cloud complexities by providing the ability to deploy database instances on multiple cloud providers. Instances that span multiple clouds can be connected using industry-standard structured query language (SQL) queries to provide an aggregated view. This distributed query offers transparent access to data wherever it resides.

The Actian Data Platform is cluster-aware for scalability. Block storage lets compute and storage scale independently. The data integration capabilities built into the Actian Data Platform with DataConnect work with popular data storage structures. For example, templates in GitHub are provided for S3 buckets, Google Drive folders, and Azure Blob storage. The Actian Data Platform makes multi-cloud management easy using a single management console. Try it by signing up for the 30-day free trial here.