PolarDB:Serverless

Background

Databases are a vital component of modern IT systems. When you create a database, you must carefully configure its resources, such as CPU, memory, storage, and connections. This configuration is intended to ensure that your business runs smoothly during both peak and off-peak hours. However, this approach often wastes resources and increases costs during periods of low demand. Conversely, during periods of high demand, the cluster may lack sufficient resources to handle the increased load. Serverless databases address this issue. They automatically scale resources up or down based on the current workload, which eliminates the need for complex resource planning and Operations and Maintenance (O&M) tasks.

The following figure compares the resource usage and specification changes of common clusters and Serverless clusters in scenarios with fluctuating workloads.

The figure shows the following differences in scenarios with fluctuating workloads:

• Common clusters: Waste resources during off-peak periods and have insufficient resources during peak periods. This can negatively impact your business.

• Serverless clusters:

  • Adjust specifications based on business demand. This reduces resource waste, improves resource utilization, and lowers overall resource usage.

  • Quickly scale cluster resources to meet business needs during peak hours. This ensures business continuity and improves system stability.

  • Break the fixed-resource payment model. The pay-as-you-go model dynamically matches resources with workloads, which can lead to significant cost savings.

  • Are optimized for high-throughput write and high-concurrency scenarios and provide elastic scaling. This makes them suitable for scenarios that involve large data volumes and fluctuating business access patterns.

  • Do not require manual configuration adjustments. This improves O&M efficiency and saves on labor and time costs.

Overview

The Serverless feature provides real-time elasticity for CPU, memory, storage, and network resources. It offers vertical resource isolation for network resources, namespaces, and storage space. It also provides on-demand billing for compute and storage resources. This lets you flexibly allocate, and independently and quickly adjust, compute and storage capacity to respond to business changes. This helps optimize costs, reduce expenses, and improve enterprise efficiency.

Presentation

Serverless cluster	Serverless feature of a cluster with defined specifications
Database Proxy The database proxy is Serverless. Its resources are independent of the compute nodes and scale elastically and automatically. By default, scaling occurs in increments of 0.5 PCU. The scaling increment is dynamically adjusted based on current PolarDB Capacity Unit (PCU) usage. A higher PCU usage results in a larger scaling increment. Compute nodes The primary node (RW node) and read-only nodes (RO nodes) are fully serverless. They scale elastically with business workloads and use shared storage in a single availability zone. When the primary node or read-only nodes scale out or in, the number of PCUs for the node increases or decreases accordingly. By default, scaling occurs in increments of 0.5 PCU. The scaling increment is dynamically adjusted based on current PCU usage. A higher PCU usage results in a larger scaling increment. You can set the elastic scaling range for a single node in PCUs. The system monitors the PCU of a compute node every second. Storage space Storage is pay-as-you-go. You do not need to specify a capacity when you purchase the cluster. The storage space scales automatically as your data volume increases. You are charged only for the storage space that you use. You can view the Database Storage Usage on the Basic Information page of the cluster. For more information, see View database storage usage. Note A serverless cluster supports a maximum of 100,000 connections and a maximum IOPS of 84,000.	Database proxy The database proxy consists of two parts: defined specifications and Serverless. The defined specifications part is the default, while the Serverless part scales elastically based on the service workload. By default, scaling occurs in increments of 0.5 PCU. The scaling increment is dynamically adjusted based on current PCU usage. A higher PCU usage results in a larger scaling increment. Compute nodes The primary node (RW node) and read-only nodes (RO nodes) include resources with defined specifications and serverless resources. The resources with defined specifications do not scale, whereas the serverless resources scale based on workloads. When the primary node or read-only nodes scale out or in, the number of PCUs for the node increases or decreases accordingly. By default, scaling occurs in increments of 0.5 PCU. The scaling increment is dynamically adjusted based on current PCU usage. A higher PCU usage results in a larger scaling increment. You can set the elastic scaling range for a single node in PCUs. The system monitors the PCU of a compute node every second. Storage space The storage of the cluster with defined specifications is used. For more information, see Storage space. Note After you enable the serverless feature for a cluster with defined specifications, the maximum number of connections and the maximum IOPS for the cluster are proportional to the value of the serverless Maximum Resources for a Single Node parameter.

Benefits

Serverless dynamically scales cluster resources in seconds based on the business workload. Its core benefits are as follows:

• High availability

  The multi-node architecture ensures the high availability and stability of Serverless clusters.

• High elasticity

  • Wide scaling range: Supports automatic vertical and horizontal scaling.

  • Scaling within seconds: Handles sudden workload spikes. Detection is completed in 5 seconds and a scale-out is completed in 1 second. When the workload decreases, cluster resources are automatically released in tiers.

• Strong data consistency

  Supports global consistency to ensure strong data consistency within the cluster. Data written to the cluster is immediately readable on read-only nodes. The performance is nearly identical to that of weak consistency.

  Note

  The global consistency feature is disabled by default. You can manually enable the feature for the cluster endpoint. For more information, see global consistency.

• Cost-effectiveness

  Serverless clusters are billed in PCUs on a pay-as-you-go basis. This can reduce your costs by up to 80%.

• Fully managed

  The Alibaba Cloud team handles all O&M work, such as version upgrades, system deployments, scaling, and alert processing. These operations run in the background without affecting your services and ensure continuous availability. This provides a fully managed experience, allowing you to focus on your business.

Scenarios

Prerequisites

Limits

Billing

• Serverless clusters

  The fees include compute node fees, storage fees, backup storage fees (charged only for usage that exceeds the free quota), and SQL Explorer fees (optional). For more information, see Serverless billing.

  Note

  Alibaba Cloud provides a 50% launch discount for serverless clusters. For more information, see Launch of the PolarDB for PostgreSQL serverless cluster.

• Serverless-enabled clusters with defined specifications

  The fees include charges for the cluster with defined specifications and for the serverless feature. For billing information about the cluster, see Billable items. For billing information about the serverless feature, see Serverless billing.