As multi-cloud becomes the new normal, the IT architecture that centrally stores and shares data and deploys applications in multiple clouds will be the optimal choice.
Continuous digital transformation encourages rapid upgrades in enterprise IT architectures. Several industries including finance, carrier, healthcare, and Internet have already started multi-cloud deployment and gained insights into best practices.
- In cloud evolution, having a combination of multiple public and private clouds has become the best choice
Cloud transformation has been accepted by a wide range of industries. It helps large enterprises take IT construction to a whole new level. It also enables small- and medium-sized enterprises to implement digital management at lower costs, instead of being overwhelmed by complex IT architecture construction.
As cloud evolution practices evolve, enterprises are keen to find the best construction mode. Initially, many enterprises preferred a single-cloud platform. However, as more types of services are deployed, the architecture of the single-cloud platform faces an increasing number of problems. On the one hand, different cloud vendors have their own strengths. On the other hand, the use of a single cloud provider increases enterprises’ concerns about data security. Therefore, going multi-cloud, which includes multiple public and private clouds, has become an important strategy for enterprises to consider in cloud evolution. Enterprises can deploy public clouds or build their own data centers to meet workload requirements.
The public cloud appeals to IT decision makers because it offers lower costs and reduced risks, as well as good elasticity. Innovative and emerging services can be migrated from on-premises systems to public clouds in order to take full advantage of the benefits offered by public cloud vendors.
In addition, local public clouds meet destination countries’ data residency and regulatory requirements, helping multinational enterprises further expand their overseas business. However, the use of multiple public clouds exposes enterprise data to public environments, which may not meet data management, control, and privacy requirements.
In contrast, enterprise-built data centers allow for independent data management, control, and operation, prevent core services from being locked in by vendors, and fully use the capabilities of enterprise-built platforms, helping enterprises take the initiative in cloud construction. Therefore, a combination of multiple public and private clouds has become the best choice for enterprises in cloud evolution.
Different public and private cloud vendors have their own advantages in IaaS, PaaS, and SaaS. To meet workload requirements, many enterprises are trying to deploy different applications on suitable cloud platforms in order to procure an optimal combination of IT technologies. In addition, global economic uncertainty threatens enterprises’ single-supplier strategy. The strategy of using multiple cloud suppliers can help enterprises avoid vendor lock-in and build a more robust IT platform. Statistics show that 89% of global enterprises have strategic plans for multi-cloud deployment. Enterprises are actively embracing multi-cloud construction and relying on cloud management services to improve the multi-cloud experience and efficiency.
- Enterprises are making efforts to remove bottlenecks in data sharing, data mobility, and unified data management across clouds
After an enterprise goes multi-cloud, data is scattered on different cloud platforms. Applications cannot invoke data across clouds. As a result, multiple data silos are created, preventing the enterprise from constructing a datacentric digital architecture. Therefore, some enterprises are seeking a new multi-cloud data architecture that centrally stores and shares data and deploys applications in multiple clouds. To share data more efficiently, acceleration engines for diverse applications are also being built on the storage side to accelerate cloud-based application deployment.
Based on application types and data security needs, enterprises attempt to deploy different applications and data in enterprise data centers or public clouds. This requires enterprises to implement unified data management and mobility across clouds for optimal efficiency in data storage and management. Unified data management refers to unified storage resource provisioning and O&M via a holistic data view. Data mobility includes global data scheduling and hybrid-cloud DR, backup, and storage tiering. Storage vendors are also actively deploying professional storage, which is widely used in data centers, on public cloud platforms by means of softwarehardware integration or pure software, helping enterprises with smooth crosscloud evolution.(Figure: Enterprise multi-cloud IT architecture)
To continue to further digital transformation, enterprises going multi-cloud will need to continue exploring ways to maximize resource sharing across clouds and data centers, implement unified data management, and benefit from an optimal combination of cloud vendors.
What we suggest
1. Migrate innovative services that have uncertainties, emerging services like OA, and archiving data to public clouds, while retaining core services in their on-premises data centers
Innovative and emerging services require IT systems to deliver elastic scaling and on-demand resource application and release. Public clouds provide benefits like low costs and reduced risks in addition to good elasticity. Enterprises are also encouraged to trial public cloud services in fields like DR, backup, archiving and cold data retention to benefit from the elasticity and largescale sharing advantages of public clouds. However, when the amount of service data reaches a certain level, the cost of migrating the data to a public cloud is much higher than that of building a data center. In addition, data tiering is also a major requirement of enterprises. For example, if data needs to be backed up half a year later due to legal requirements, the tiered storage capability will help simplify the cloud migration.
To maintain the competitive edge of core services, enterprises must develop strong capabilities in IT platform R&D. However, public clouds only provide standard IT services. Therefore, retaining core services in enterprises’ on-premises data centers and tiered data mobility can truly stimulate IT R&D and innovation, implement independent data control and operation, and prevent cloud vendor lock-in, helping enterprises take the initiative in building digital competitiveness.
2. In multi-cloud construction, enterprises are advised to use the IT architecture that centrally stores and shares data and deploys applications in multiple clouds, and plan a unified data management platform across clouds to maximize data sharing
When selecting storage to build a unified data management platform across clouds, consider the following points:
- Leverage the technical advantages of multiple cloud vendors for faster application development and select an IT architecture with application storage decoupling for data sharing across clouds.
- Promote storage vendors to deploy professional storage on public cloud platforms by means of software-hardware integration or pure software, thereby helping with smooth cross-cloud evolution.
- Plan a unified data management platform across clouds to implement a holistic data view, global data scheduling, and hybrid-cloud DR/ backup/tierin. Therefore, data is stored by service level on suitable storage tiers with efficient data mobility, which also achieves the reuse of legacy systems and maximizes the value of legacy data.
3. Enterprises and storage vendors actively collaborate on innovation projects based on the latest cloud technologies like containers
As databases, AI-powered analytics, middleware, and content services develop, enterprises need to work on shortening their applications’ time to market (TTM) and reconstructing applications into microservices. Containers are the best carrier of the microservice architecture and have been adopted by most industries, especially the finance and Internet industries.
Containers are evolving from stateless to stateful applications. Therefore, persistent storage is needed to carry containers. In addition, storage systems need to work with container management platforms to enable large-scale provisioning of containerized applications. Disaster recovery of containerized applications also requires close cooperation between container management platforms and storage systems. Therefore, enterprises and storage vendors should actively embrace the latest cloud technologies (such as containers) and carry out joint innovation projects to fully unlock the value of multiple clouds.
Most SaaS and PaaS in the cloud are implemented through container capabilities. When migrating data to on-premises data centers, storage devices are required to provide the same capabilities as the cloud to ensure continuous IT evolution in the future.
Learn more about Huawei’s Data Storage solutions.
Article Source: HuaWei
Article Source: HuaWei
Disclaimer: Any views and/or opinions expressed in this post by individual authors or contributors are their personal views and/or opinions and do not necessarily reflect the views and/or opinions of Huawei Technologies.