Trends for Databases in the Cloud Era: Distributed, Multi-Cloud, Multi-Model
The world has firmly entered the cloud era, so where will cloud databases go from here? Here are some of our insights.
1. Cloud Database Technology Enhances Cost Efficiency for Enterprises
Today, companies face unprecedented challenges and opportunities, with two primary goals in focus: enhancing business efficiency and reducing costs. While the benefits of cloud for efficiency are widely recognized, its potential to lower costs is equally valuable. Here’s how:
(i) Resource Pooling to Reduce Hardware Costs
Most companies depend on IT infrastructure, yet computer utilization remains low, often with CPUs operating at single-digit usage levels. The cloud aggregates business and technical resources into a shared pool, maximizing the use of hardware. By doubling utilization, businesses can essentially get twice the computing power for the price of one. This advantage drives significant value for both cloud providers and users.
(ii) Lower Manpower Costs with Cloud Services
Providing top-tier database services often involves on-site assessments and system maintenance, requiring considerable travel time. Cloud-enabled online services can streamline these processes by moving to the cloud, saving time and boosting efficiency without the physical presence required.
2. Distributed, Multi-Cloud, and Multi-Model: The Future of Databases in the Cloud Era
The rapid advancement of cloud technologies and services creates new opportunities in the cloud database space. Notably, two cloud-native providers rank among the top global database platforms. However, the challenges of standalone, single-model, and single-cloud still plague cloud database development.
(i) Standalone Deployment Wastes Hardware Resources
Despite over five decades of advancements, most databases today still operate on standalone systems, which struggle to handle large data volumes efficiently. Physical CPUs are tied to individual machines, leading to resource allocation issues like memory fragmentation. For example, in a scenario where 5 machines have independently deployed standalone databases, requesting the use of 5 or 6 CPU cores on one machine might be impossible, even if the system has the resources. Cloud-based distributed systems address this issue by enabling flexible resource allocation—pooling CPU cores across machines to optimize utilization and lower costs.
(ii) Single-Model Processing is Time and Resource Intensive
In the era of big data, databases contain vast amounts of data requiring specialized systems for different functions, such as transactional processing, analytical processing, and document management. Businesses often expend significant time, effort, and cost to synchronize data between these various systems, akin to syncing data between transactional databases, analytical databases, and big data platforms. Just as a smartphone integrates features that used to require separate devices, multi-model integrated databases reduce the need for disparate systems.
(iii) Single-Cloud Deployment Increases Business Risk
Many cloud providers offer data services, but most databases are restricted to their cloud platforms, limiting flexibility for business users requiring local region cloud services or cross-cloud setups for disaster recovery. Deploying the same database system across multiple clouds can introduce challenges like MySQL compatibility, which can increase business risks.
Cloud has become a hallmark of this era as it brings convenience, ease of use, and efficiency, lowering costs through resource pooling. However, the evolution of cloud databases still faces restrictions.
While standalone, single-model, and single-cloud setups have significant limitations in most business applications, their evolution mirrors the broader history of technological change. The move from standalone databases to distributed systems, from single to multi-model databases, and from single to multi-cloud architectures represents an inevitable trend in the evolution of data management.
