Help Center/ GeminiDB/ GeminiDB DynamoDB-Compatible API/ Working with GeminiDB DynamoDB-Compatible API/ GeminiDB Serverless/ Introduction to GeminiDB Serverless
Updated on 2026-01-30 GMT+08:00
View PDF
Share

Introduction to GeminiDB Serverless

In traditional fixed-capacity database architectures, you have to pre-define instance specifications and throughput limits. This static capacity model often results in overload during peak hours .and wastes resources during off-peak hours, falling short of the demand for highly scalable and cost-effective applications.

To address these issues, GeminiDB Serverless comes in. It supports pay-per-use billing, autoscaling, and worry-free O&M. You do not need to pre-configure storage or compute resources. They are automatically scaled and billed based on actual resource usage. You can focus on application logic without worrying about infrastructure management.

Usage Notes

This function is now in OBT. To use it, choose Service Tickets > Create Service Ticket in the upper right corner of the console and contact the customer service.

Billing

You can use this function for free during OBT and will be billed after GA. For more information, be on the lookout for notices on Huawei Cloud.

You need to pay for creating a VPC endpoint. For details, see VPC Endpoint Billing.

Highlights

With decoupled storage and compute, GeminiDB Serverless provides the following capabilities:

  1. Autoscaling
    • Automatic compute scale-out: Resources are dynamically scaled based on the actual workloads without manual intervention.
    • Automatic storage scale-up: Storage is automatically scaled up as data volumes increase. You do not need to pre-define the maximum storage.
    • Response in seconds: Resources can be scaled up in seconds to ensure service stability during traffic spikes (such as promotions or crawlers).
  2. Pay-per-Use
    • Compute: You are billed only when requests are processed. If no requests are processed, Serverless is free of charge or has low baseline costs.
    • Storage: Storage is billed per GB per hour to avoid resource wastes and enable fine-grained cost control.

    This billing mode significantly reduces user costs and is especially suitable for scenarios with considerable workload fluctuations and intermittent usage, such as IoT devices, development and test environments, and SaaS applications.

  3. Resource sharing and isolation across tenants
    • Shared resource pools with logical isolation and reservation ensure stable performance across users.
    • Multiple instances can be deployed in the same physical cluster, improving resource utilization while ensuring security and efficiency.

Scenarios

  1. IoT device data access: Millions of devices can intermittently report data. You are billed only for what you use. Resources are automatically scaled to accommodate workload fluctuations.
  2. Caching/Shopping cart for e-commerce promotions: Resources are automatically scaled in after activities to reduce costs and ensure system stability.
  3. Real-time social/game data: Messages and ranking data are read and written with low latency. GeminiDB Serverless responds in milliseconds, ensuring smooth experience.
  4. DevTest/CI environment: Resources are billed by second on the pay-per-use basis and are deleted once consumed, eliminating O&M overhead and enhancing development efficiency.
  5. Quick verification of innovative projects: You do not need to buy resources in advance. They can be deployed within minutes.
Parent Topic: GeminiDB Serverless