Skip to main content
Version: latest

CnosDB vs InfluxDB

CnosDB 2.4.1 vs InfluxDB 1.8.10

Comparative performance testing between CnosDB 2.4.1 and InfluxDB 1.8.10 has been conducted, below are the test conclusions and test details.

Conclusion

Under the same data conditions, test results show that the query performance of CnosDB and InfluxDB have their own advantages and disadvantages.

Pre-test

1. Test Environment Preparedness

CPU:64 CPUs x Intel(R) Xeon(R) Gold 5218 CPU @ 2.30GHz

Memory: 256 GB

Disk: SSD NVMe Protocol

2. Test case preparation

  1. Install the corresponding machine's db environment, go environment, etc. in advance to ensure that you can connect properly.

  2. Install CnosDB:

    Refer to the deployment document: Install CnosDB

  3. Test InfluxDB writes

    Reference: [InfluxDB 1.8.10] (https://github.com/influxdata/influxdb)

3) Configuration Check & Modify

CnosDB and InfluxDB only modified the storage folder paths for Data, Wal, and Meta, while keeping the rest as default. This will not be repeated here.

4. Dataset preparation

UsageDetermine the PRNG-seedNumber of devices to generateStart timestampEnd timestampInterval between readings per deviceTarget databaseData SizeRows
iot1231002020-01-01T00:00:00Z2021-01-01T00:00:00Z6.3sCnosDB/InfluxDB201G450,721,871

Test Results

SQLCnosDB 2.4.1InfluxDB 1.8.10
Count: select count(elevation) from readings2.88s2.09s
Aggregate: select count(latitude), max(latitude), min(latitude) from readings3.2s4.95s
Tag filter: select count(latitude), max(latitude), min(latitude) from readings where fleet = 'East'1.75s2.67s
Tag group: select count(latitude), max(latitude), min(latitude) from readings group by driver6.11s6.15s