A DTS Serverless instance is an instance of the Data Transmission Service (DTS) with elastic resource specifications. It adapts to changing business needs by automatically adjusting resources as your workload scales. This helps avoid resource waste and reduces operations and maintenance (O&M) costs.
Background information
Serverless is a dynamic billing method. It adjusts resources in minute-level increments based on workload and bills you in real time, generating one bill per hour. You pay only for what you use, which saves significant costs. Instances purchased with this billing method are called Serverless instances.
A Serverless instance dynamically adjusts its resource specifications based on factors such as RPS (records per second), CPU, memory utilization, and network. The adjusted resource specifications are expressed in DUs (DTS Units). Sixty seconds after the DU count is adjusted, the system checks if the current resource specifications meet the load requirements.
In scenarios with highly variable data transfer volumes, resource usage and specification changes differ between standard and Serverless instances, as shown in the following figure:
The figure shows that in scenarios with high business fluctuations:
Standard instance: Wastes significant resources during low-demand periods and lacks sufficient resources during peak periods, which can impact service.
Serverless instance: Dynamically adjusts resource specifications to match workload demand. It fully meets business needs during both low-demand and peak periods, ensuring uninterrupted service.
Advantages
High elasticity
Resource specifications adjust in real time (within minutes) as data transfer volumes change. This improves resource utilization and reduces overall resource consumption.
Fully managed
No manual upgrades or downgrades are needed, which improves O&M efficiency and lowers O&M costs for O&M staff and developers.
Low cost
It eliminates fixed-resource pricing. It enables true pay-as-you-go billing, with resources scaling dynamically based on actual load. This saves significant costs.
Scenarios
Your data transfer volume is large initially but stabilizes later.
Your data transfer follows clear peak and off-peak patterns, and you want to reduce continuous running costs.
You have existing data synchronization tasks and want to either purchase new Serverless instances or convert standard sync instances to Serverless instances.
Your data transfer volume is unpredictable.
You are an individual developer.
Limits
The following Serverless instance types are undergoing grayscale testing and are available only to select users.
Cross-region or cross-border replication: Supports only cross-region replication between Southwest 1 (Chengdu) and South China 3 (Guangzhou). All other regions are not supported.
Two-way synchronization.
The Database Type of the source database is Tair/Redis, PolarDB-X 1.0, PolarDB-X 2.0, or Sharded Cluster MongoDB.
The Database Type of the destination database is Tair/Redis.
The minimum DU values can be 1, 2, 4, 8, 16, or 32. The maximum DU values can be 2, 4, 8, 16, or 32.
Cost details
For more information, see Serverless instance billing method.
Supported regions
Serverless instances are available in the following regions: China (Hangzhou), China (Shanghai), China (Qingdao), China (Beijing), China (Zhangjiakou), China (Ulanqab), China (Shenzhen), China (Guangzhou), China (Wuhan - Local Region), China (Chengdu), China (Hong Kong), Singapore, Malaysia (Kuala Lumpur), Indonesia (Jakarta), Japan (Tokyo), and Germany (Frankfurt).
Performance details
The performance of a Serverless task depends on the number of DUs, as shown in the following table. Actual performance may vary due to network conditions, source and destination database performance, latency, and other factors.
DU count | Performance ceiling (RPS) reference |
1 | 200 |
2 | 1,000 |
4 | 3,000 |
8 | 9,000 |
16 | 17,000 |
32 | 35,000 |
Procedures
FAQ
Q: Can I create Serverless instances for data migration or change tracking tasks?
A: No, only data synchronization tasks support Serverless instances.