Elasticsearch Service
- Release Notes and Announcements
- Product Announcements
- Security Announcement
- Product Introduction
- Performance
- ES Kernel Enhancement
- Getting Started
- ES Serverless Guide
- Quick Start
- Index Management
- Data Application Guide
- Elasticsearch Guide
- Cluster Scaling
- Cluster Configuration
- Plugin Configuration
- Monitoring and Alarming
- Log Query
- Practical Tutorial
- Data Migration and Sync
- Use Case Construction
- Index Configuration
- API Documentation
- Instance APIs
- Making API Requests
- FAQs
Elasticsearch Service Serverless Pricing
Last updated: 2024-08-15 10:52:24
Elasticsearch Service (ES) Serverless is billed based on the actual usage of traffic resources, storage resources, and service requests generated during reads and writes. Currently, it supports the pay-as-you-go mode, which means billing is done hourly based on the actual usage of each billing item for the previous hour.
Pay-as-You-Go
Billing Item
Billing Item | Billing Item Description | Billing Description |
Traffic calculation | It indicates the traffic size generated during data write, indexing, and segmentation. The size of the traffic depends on the fields in the data and the field length. | Calculation formula: (0.1 + Proportion of Fields Indexed) * Daily Increment of Raw Data Size. The proportion of fields indexed is generally assumed to be 100%. For example, if the size of raw data written daily is 1 GB and indexes are built for all fields, the size of the calculated traffic would be 1.1 GB. |
Data storage | Data storage size is related to the raw data size, data inflation generated during indexing, and storage duration. | Calculation formula: (0.12 + Proportion of Fields Indexed) * Daily Increment of Raw Data Size * storage duration. The proportion of fields indexed is generally assumed to be 100%. For example, if the raw data size written daily is 1 GB (data storage of 3 days) and indexes are built for all fields, then the data storage size will stabilize at 3.36 GB after the 4th day. |
API calling | Calculate based on the number of times ES Serverless API is actually called. The number of times that APIs are called during the write process can be reduced through batch write. | / |
Pricing Description
Region | Billing Item | ||
| Traffic Calculation (USD/GB) | Data Storage (USD/GB/Day) | API Calling (USD/million times) |
Beijing/Shanghai/Guangzhou /Nanjing | 0.051 | 0.0021 | 0.017 |
Hong Kong (China) | 0.054 | 0.0021 | 0.019 |
Singapore | 0.054 | 0.0021 | 0.019 |
Tokyo | 0.054 | 0.0021 | 0.019 |
Virginia | 0.060 | 0.0023 | 0.017 |
Fees Calculation Cases
Assume that the access log of your App averages 100 bytes per item, with a total of 100 million logs per day (approximately 9.3 GB of raw log data per day) and data storage for 7 days. Indexes are built for all fields, and App logs are analyzed by using ES Serverless in Guangzhou. In this case, the data storage size will stabilize at 72.9 GB after the 8th day. The cost for one day will be approximately USD 0.675. The specific details are shown in the following table:
Billing Item | Description | Usage Estimation | Unit Price | Cost Estimation (USD) |
Traffic calculation | Indexing all fields, indexed fields account for 100%. | (0.1 + 100%)* 9.3 = 10.23(GB) | 0.0510 USD/GB | 0.52170 |
Data storage | Indexing all fields, indexed fields account for 100%. | (0.12 + 100%)* 9.3 * 7 = 72.90(GB) | 0.0021 USD/GB/day | 0.15309 |
API calling | Bulk size is 5,000 documents per times | 100 million documents/bulk size = 20,000 times | 0.0170 USD/million times | 0.00034 |
Note
The proportion of fields indexed is mainly used to estimate the traffic and storage consumption generated during the indexing process. For example, if you have 10 fields of the same type and equal length and you index 5 of them, the proportion of fields indexed is 50%. This calculation method is for reference only, and the actual resource consumption depends on your field values and lengths. When the proportion of fields indexed is less than 40%, it is calculated by default as 40%.
Was this page helpful?
You can also Contact Sales or Submit a Ticket for help.
Yes
No