On February 21, TDengine, an open-source, high-performance, and cloud-native time-series data platform, officially released the TDengine 3.0 performance comparison analysis report, which compared TDengine with other popular time-series databases in the market ( Time Series Database, TSDB) product performance differences. The report aims to verify the performance advantages and cost control level brought by TDengine's unique architecture based on time-series data scenarios. All tests are completed using public data under standardized conditions.
The performance benchmarking platform chosen by TDengine is Time Series Benchmark Suite (TSBS). TSBS is a time-series database performance benchmark evaluation platform created by Timescale and open source, which integrates time-series data generation, data writing, query processing, automatic result summary statistics and other functions in a variety of application scenarios. Due to its open-source nature, the platform has been used by many leading time-series database vendors around the world since 2018, becoming the most widely used platform for time-series database performance testing. The newly released performance test report of TDengine uses the DevOps scenario in the TSBS platform as the basic data set, runs TDengine 3.0, TimescaleDB 2.6 and InfluxDB 1.8 in the same AWS cloud environment, conducts comparative tests from five dimensions and outputs the results.
"Cost reduction and efficiency increase have always been an important goal of the sustainable development of enterprises, and as the magnitude of time-series data in the business continues to increase significantly, the performance of the selected database will directly affect the response time and operating costs of the system." Founder of TDengine Tao Jianhui, the core developer of Renhe, said, "With the release of the TDengine 3.0 TSBS performance test report, we are more confident than ever that no matter how the data scale increases or what stage the enterprise is in, TDengine can provide it at the lowest cost. provide the required performance."
Highlights of the test report include:
Data writing comparison : In all five scenarios, the writing performance of TDengine is better than that of TimescaleDB and InfluxDB. The write performance is up to 6.7 times that of TimescaleDB and 10.6 times that of InfluxDB. In addition, TDengine consumes minimal computing (CPU) resources and disk IO overhead during writing.
Data query comparison : For most query types, TDengine's performance is better than InfluxDB and TimescaleDB, and it shows a huge advantage in Complex queries type queries - TDengine's Complex queries query performance is up to 37 times that of InfluxDB, TimescaleDB's 28.6 times.
Data storage comparison : Due to the efficient data storage and compression mechanism, in some scenarios, the disk space occupied by TDengine is much lower than that of InfluxDB and TimescaleDB when storing data of the same size. The report shows that as the size of the data set grows, TDengine has more obvious advantages in data storage, which directly shows that TDengine is more suitable for the storage of time-series big data. Under the same data scale, the data scale of TimescaleDB is up to 26.9 times that of TDengine, and the disk usage of InfluxDB is up to 4.5 times that of TDengine.
Resource consumption comparison: From the perspective of overall CPU overhead, TDengine not only takes less time to complete all queries than TimescaleDB and InfluxDB, but also consumes much less CPU computing resources than TimescaleDB and InfluxDB as a whole. During the entire query process, TDengine memory is also maintained in a relatively stable state.
For details on the test report execution results and steps to reproduce, see the full report: https://www.taosdata.com/performance-comparison-influxdb-and-timescaledb-vs-tdengine .
About TDengine
TDengine is an open-source, high-performance, cloud-native time-series database completely independently developed by Taos Data, focusing on the storage, query, analysis and calculation of time-series space big data. The number of TDengine user instances running worldwide exceeds 200k, with an average of new Hundreds of new deployments have been added, with users in more than 50 countries/regions around the world, and it has been widely used in the Internet of Things, Internet of Vehicles, Industrial Internet, IT operation and maintenance and other fields. In July 2019, TDengine was open-sourced on GitHub. At present, the number of stars on GitHub has reached 20.8k, and it has repeatedly topped the GitHub global trend list.
In August 2022, TDengine launched version 3.0, which has truly become a cloud-native time-series database, supports data collection by 1 billion devices, 100 nodes, supports the separation of storage and computing, and solves the problem of high cardinality that plagues the development of time-series databases; The storage engine and query engine have been optimized and upgraded, and a brand-new streaming computing engine has been created, which eliminates the need to integrate Kafka, Redis, Spark, Flink and other software. While the performance is improved, the complexity of the system architecture is also greatly reduced. In September of the same year, TDengine Cloud went online in overseas markets and simultaneously supported the three major public cloud platforms of Microsoft Azure, AWS, and Google Cloud. The TDengine PI connector was also successfully launched shortly thereafter. Get all the benefits a modern cloud platform has to offer with ease. Click to enter www.taosdata.com for more information.
If you want to know more about the specific details of TDengine Database , you are welcome to view the relevant source code on GitHub .