More about Multicloud Storage
- Multicloud Kubernetes: Centralizing Multicloud Management
- Multicloud Storage: Everything You Need to Know
- Multicloud Architecture: Partitioned, Cloud Burst and DR
- Google Hybrid Cloud with Anthos and Cloud Volumes ONTAP
- Hybrid Deployment with Google Anthos: An Intro
- Multicloud & Hybrid Architectures: Benefits and Challenges
A multicloud deployment needs to address the unique and often interdependent needs of three key layers: infrastructure, operations, and applications. In order to avoid post-deployment disconnects and conflicts, a carefully crafted plan should be put in place that aligns all of these layers with the objectives of the organization’s strategy for multicloud storage.
This blog post reviews the main considerations and important aspects of a multicloud deployment plan, and shows how Cloud Volumes ONTAP offers practical solutions to support them.
Why Use a Multicloud Deployment?
What is Multicloud?
Multicloud refers to an IT deployment that relies on cloud services provided by multiple public cloud vendors. A multicloud deployment can be either distributed, with each app and its data deployed modularly across multiple clouds, or redundant, with backups or complete standby systems replicated in multiple clouds.
There are many benefits to using multicloud architecture. All of the many cloud usage reports published throughout 2019 point to extensive enterprise adoption of a multicloud architecture, including hybrid multicloud architectures, which combine multiple clouds with on-prem systems. Gartner, for example, found that 81% of the organizations surveyed are working with two or more public cloud providers. Similarly, Turbonomic’s 2019 State of Multicloud Report shows that just under 80% of organizations that use the public cloud are deployed across more than one vendor, with the median number being two.
The cloud providers themselves are acknowledging the shift to multicloud environments and providing their own tools to make multicloud management easier. Google Anthos enables users to orchestrate workloads on-prem and in multiple clouds. Not to be left behind, Microsoft announced Azure Arc in November 2019, a technology suite that allows Azure cloud customers to deploy and manage Azure services on any cloud infrastructure, including AWS and Google Cloud.
The main benefits that are driving the growth of multicloud deployments are:
- Optimized ROI: Finding the best cloud IaaS resource price points across multiple vendors.
- Enhanced application and operational efficiency: Mixing and matching best-of-breed SaaS, PaaS, and fully managed solutions.
- Improved latency by taking advantage of geographically distributed data centers.
- Better availability and business continuity: Redundant multicloud deployments support high availability and also minimize Single Point of Failure (SPOF) disruptions.
- Business agility: Avoiding vendor lock-in.
Key Considerations When Planning a Multicloud Deployment (and How Cloud Volumes ONTAP Helps to Address Them)
What is Cloud Volumes ONTAP?
Cloud Volumes ONTAP is a data management solution for enterprise workloads. Running as a virtual machine on all the leading public cloud platforms (AWS, Azure, and Google Cloud), Cloud Volumes ONTAP’s powerful data protection and storage efficiency features are available seamlessly across the enterprise’s entire multicloud/hybrid environment. The enterprise benefits from an end-to-end data fabric that is managed centrally from a single-pane management portal (NetApp Cloud Manager).
Five key infrastructure considerations to be addressed by a multicloud deployment plan are:
1. Specify the target infrastructure based on the current and future needs of all relevant stakeholders including: the different lines of business, corporate and administrative backend platforms, and all application lifecycle requirements, from development to testing, staging and production.
Cloud Volumes ONTAP is a versatile data management solution that supports a wide range of business verticals, lines of business, and use cases including databases such as SQL in the cloud, file shares, backup and archives, disaster recovery, and DevOps.
2. Take into account the impact of advanced infrastructure technologies such as virtualization, software-defined infrastructure, flash and solid-state storage arrays, and converged/hyper-converged infrastructure.
✔ Solution: Cloud Volumes ONTAP supports all these technologies across even the most complex multicloud and hybrid environments.
3. Plan for the data format conversions that will be required when you move data between on-premises and public cloud, and from one cloud provider to another.
✔ Solution: Cloud Volumes ONTAP supports multiple protocols and data formats including NFS, SMB and SAN/iSCSI and automatically takes care of conversions as data moves across the enterprise data fabric.
4. Support infrastructure self-provisioning as much as possible, such as approved service catalogs and Infrastructure as Code (IAC) templates. Bear in mind, however, that the cloud providers’ IAC tools such as Azure Templates, AWS Cloud Formation, or Google’s GCP Deployment Manager, are vendor-specific and difficult to manage in a multicloud environment.
✔ Solution: Via Cloud Manager’s API, an enterprise can use any IAC, DevOps, or scripting tool platform, including Ansible or Terraform, to seamlessly standardize and automate deployment and configuration of Cloud Volumes ONTAP operations across multicloud and hybrid infrastructures.
5. Ensure that data being stored on Kubernetes persistent volumes is properly managed and protected. Containers have become popular because the code they run always executes in the same way regardless of the infrastructure on which they are deployed. This is a big advantage in multicloud and hybrid environments. However, stateful containerized applications require persistent storage with independent lifecycles—a requirement that is met by the dynamic provisioning of persistent volumes. Protection of the data stored on these volumes is not handled by Kubernetes but by the third-party services that provision and store persistent volumes.
✔ Solution: Using Trident and Cloud Volumes ONTAP, Kubernetes storage can be dynamically provisioned for stateful data sets and gain the benefits of flexible data management and enterprise-grade data protection.
Your multicloud deployment plan must take into account many operational issues, including the ones outlined below.
1. The deployment will change the IT landscape and new roles may need to be created. Some examples: Business Relationship Managers to establish and maintain a close alignment between business needs and IT services; Service Portfolio Managers to manage the catalog of services and their life cycles; and Demand Managers to structure infrastructure and human resources so that service demands are met.
✔ Solution: Cloud Manager’s authentication and authorization features ensure that the growing cadre of multicloud managers has granular access to the resources and services required to fulfill their roles, without compromising multicloud security.
2. Decouple the cloud provider’s storage management layer from storage volume requirements.
✔ Solution: Command Manager provides a central, single-pane storage management layer that allows workloads and applications to be cloud-agnostic. With storage configuration handled transparently by Cloud Volumes ONTAP, storage volumes are seamlessly interoperable across clouds.
3. A cost management process must be put in place to handle current and future right-sizing, with cost management being considered the #1 challenge of multicloud deployments.
✔ Solution: Cloud Volumes ONTAP’s wide array of storage efficiencies that lower multicloud storage footprint and costs: thin provisioning for right-sizing storage through dynamic allocation, data compression and deduplication, automatic tiering of infrequently-used data to lower-cost object storage, automated data storage tiering, incremental snapshots based on NetApp Snapshot™ technology, and zero-footprint data cloning.
4. The ability to easily move data across clouds when needed.
✔ Solution: Using NetApp SnapMirror® for cost-effective data replication and synchronization, moving data across multicloud (and hybrid) architectures is a simple drag-and-drop operation in Cloud Manager.
Cloud Manager view - multi-cloud environment and data replication relationships
5. Institute a pricing and billing model that supports transparency for the full range of multicloud end-users.
✔ Solution: Cloud Manager lets organizations deploy, manage and oversee the entire Data Fabric from a centralized portal. Cloud Manager monitors data usage across clouds and shows how much was saved by using Cloud Volumes ONTAP’s storage efficiency and data tiering features.
1. Consider carefully which applications and workloads are best suited to which cloud platform. Some of the factors that should be taken into account are: the geographic locations of the provider’s data centers, which can be important for both latency and compliance reasons; the availability (and price points) of specialized compute instances for memory-bound or compute-bound workloads; and the ease of integrating the cloud provider’s resources and services with other cloud environments.
✔ Solution: Application and workload requirements can change over time. Because Cloud Volumes ONTAP abstracts the storage layer from the application layer, applications can be easily transferred from one cloud to another with no need to change code or other special adjustments. The applications will run the same since Cloud Volumes ONTAP is fully compatible with all leading data types, formats and versions, such as NFSv3, NFSv4, and SMB.
2. Development stacks should be standardized and coordinated as best as possible across clouds.
✔ Solution: Cloud Volumes ONTAP leverages NetApp FlexClone for instant, writable data cloning of storage volumes for dev/test purposes. SnapMirror can then be used to deploy these clones anywhere within the multicloud environment.
3. Data security is the #2 challenge of multicloud deployments.
✔ Solution: In addition to its powerful authentication and authorization features, which are key to securing data, Cloud Volumes ONTAP ensures that workload data and files are encrypted both at-rest and in-transit as they move freely around the multicloud Data Fabric. NetApp manages the data encryption while the customer retains the encryption keys. Read more about Cloud Volumes ONTAP data security features here.
4. Protecting data against loss or corruption due to natural disasters, system failures or human error is a core requirement of any multicloud deployment plan.
✔ Solution: Cloud Volumes ONTAP supports high availability (HA) across multiple zones and achieving the most stringent business continuity objectives (RPO=0, RTO<60). In addition, Cloud Volumes ONTAP Snapshots are available in all clouds, forming the backbone of backup and disaster recovery capabilities— including failover and failback/recovery— across all
A Final Note
Multicloud deployment is often transformational for an enterprise. As with any large-scale, strategic endeavor, deploying the multicloud roadmap should be done in an agile manner. As in product development—where time-to-market is optimized by initially releasing a basic version of a product, which is later improved upon—an enterprise adopting a multicloud architecture should define and deploy an essential set of services before ultimately achieving the full optimal multicloud deployment.
Cloud Volumes ONTAP supports the infrastructure, operational and application requirements of a multicloud deployment throughout its lifecycle, creating an enterprise Data Fabric across complex hybrid and multicloud architectures. Cloud Manager, a single-pane centralized console, abstracts the storage management layer from storage infrastructure, allowing the transparent movement of data and workloads across all the major cloud providers.
To experience this versatile, provider-agnostic data storage management solution first-hand, get started with a Cloud Volumes ONTAP 30-day trial on AWS, Google Cloud, or Azure.