tencent cloud

TencentDB for CTSDB

Release Notes and Announcements
Release Notes
Announcements
Product Introduction
Overview
System Architecture
Use Cases
Key Concepts
Instance Types and Specifications
Purchase Guide
Product Pricing
Purchasing an Instance
Renewal Instructions
Overdue Payment Instructions
Refund Instructions
Getting Started
Connecting to an Instance and Writing Data
Operation Guide
CAM
Managing an Instance
Automatic Backup
System Monitoring
Database Management
Account Management
Managing a Security Group
Public Network Access
InfluxQL Compatibility
Database Management
Schema Query
Data Type
Data Query
InfluxQL Functions
Client
SDK Reference
influx CLI Client
API Documentation
History
Introduction
API Category
Making API Requests
Instance APIs
Database APIs
Data Types
Error Codes
Practical Tutorial
Recommendations for the Data Table Definition
Using the Grafana Service
Integrating the Telegraf Service
Performance White Paper
Test Environment
Testing Tool
Write Performance Test
Query Performance Test
FAQs
Glossary
CTSDB Policy
Privacy Policy
Data Privacy and Security Agreement
Contact Us

Write Performance Test

PDF
Focus Mode
Font Size
Last updated: 2026-01-23 17:32:06

Test Objective

This document aims to measure the number of measurement points written per second when time series data is continuously written to InfluxDB under specific configurations. By systematically adjusting key parameters, the test measures and optimizes the write performance of InfluxDB, providing a performance benchmark reference for deployment in production environments.

Key Test Parameters

Parameter
Description
Tuning Suggestion
Worker (Concurrency)
Number of clients that simultaneously initiate write requests to the database
Start from a lower value (such as 4–8)
Gradually increase the value until performance no longer improves.
Use it to test the system concurrent processing capability.
Batch (Batch Size)
Number of data rows written in each batch, with a default of 10,000
Impact of data volume on a single write operation:
If the volume is too small, the network overhead will be increased.
If the volume is too large, the memory pressure will be increased.

Test Configuration Example

cat data/influx-data-cpu-10000-30day.gz | gunzip | ./tsbs_load_influx \\
--urls=http://XX.XX.XX.XX:8086 \\ # The InfluxDB cluster address.
--db-name=test \\ # The target database name.
--workers=10 \\ # Set the number of concurrent workers to 10.
--user=ctsdbi-mjhiaefp \\ # The username for authentication.
--password=Xstor@2025 \\ # The password for authentication.
--do-create-db=false # Assume that a database already exists and do not create a database.


Test Results


Test Instance Specifications
Test Concurrency
Write Performance (Metrics/s)
16-core 64 GB*3
1
6732619.16
5
28103638.38
10
35870680.17
20
40126082.02
40
41324262.38

Help and Support

Was this page helpful?

Help us improve! Rate your documentation experience in 5 mins.

Feedback