Test Results for Large Data Sets
Last updated:2025-12-25 10:46:59
This document introduces the performance test results for general TDSQL-C for MySQL clusters in read-only, mixed read-write, and write-only scenarios where the data set characteristic is large data set.
Definition of the Large Data Set
The large data set scenarios refer to the situations where all data cannot be fully stored in the cache (with capacity being twice the memory size), and disk read and write operations are required during queries to update the cache.
Test Scenarios and Read Types for Large Data Sets
Data Set Characteristic | Test Scenario | Read Type |
Large data set | Read-only | range select |
| | point select |
| Mixed read-write | range select |
| | point select |
| Write-only | - |
Test Results for Large Data Sets
Scenario 1: Read-Only (range select)
Specifications | Concurrency | Data Volume of a Single Table (table_size) | Total Number of Tables (tables) | TPS | QPS |
2 cores and 4 GB | 64 | 800000 | 100 | 986.18 | 15778.9 |
4 cores and 8 GB | 128 | 800000 | 100 | 1899.81 | 30396.95 |
8 cores and 16 GB | 128 | 800000 | 300 | 3475.23 | 55603.71 |
8 cores and 32 GB | 128 | 800000 | 300 | 3856.53 | 61704.5 |
8 cores and 64 GB | 256 | 800000 | 450 | 3981.17 | 63698.88 |
16 cores and 64 GB | 256 | 800000 | 450 | 7569.4 | 121110.61 |
16 cores and 96 GB | 256 | 800000 | 600 | 7886.19 | 126179.12 |
16 cores and 128 GB | 256 | 800000 | 600 | 9339.65 | 149434.51 |
24 cores and 144 GB | 512 | 5000000 | 300 | 9822.29 | 165156 |
32 cores and 128 GB | 128 | 5000000 | 300 | 11414.97 | 182639.83 |
32 cores and 256 GB | 128 | 5000000 | 400 | 12116.51 | 193864.33 |
64 cores and 256 GB | 1024 | 5000000 | 400 | 22663.31 | 362613.37 |
64 cores and 512 GB | 256 | 6000000 | 450 | 25839.43 | 413431.47 |
88 cores and 710 GB | 32 | 7000000 | 450 | 53595.01 | 857521.18 |
Scenario 2: Read-Only (point select)
Specifications | Concurrency | Data Volume of a Single Table (table_size) | Total Number of Tables (tables) | TPS | QPS |
2 cores and 4 GB | 64 | 800000 | 100 | 2326.38 | 27916.62 |
4 cores and 8 GB | 64 | 800000 | 100 | 4700.31 | 56403.83 |
8 cores and 16 GB | 128 | 800000 | 300 | 8612.8 | 103353.87 |
8 cores and 32 GB | 128 | 800000 | 300 | 9860.97 | 118331.87 |
8 cores and 64 GB | 256 | 800000 | 450 | 10164.74 | 121977.1 |
16 cores and 64 GB | 256 | 800000 | 450 | 18414.44 | 220973.8 |
16 cores and 96 GB | 256 | 800000 | 600 | 20026.74 | 240321.34 |
16 cores and 128 GB | 256 | 800000 | 600 | 24866.86 | 298402.91 |
24 cores and 144 GB | 512 | 5000000 | 300 | 25879.96 | 310560.14 |
32 cores and 128 GB | 128 | 5000000 | 300 | 29613.01 | 355356.85 |
32 cores and 256 GB | 1024 | 5000000 | 400 | 30943.51 | 371322.9 |
64 cores and 256 GB | 256 | 5000000 | 400 | 47441.4 | 569298.06 |
64 cores and 512 GB | 256 | 6000000 | 450 | 56501.67 | 678021.5 |
88 cores and 710 GB | 1500 | 7000000 | 450 | 71459.94 | 857521.18 |
Scenario 3: Mixed Read-Write (range select)
Specifications | Concurrency | Data Volume of a Single Table (table_size) | Total Number of Tables (tables) | TPS | QPS |
2 cores and 4 GB | 64 | 800000 | 100 | 833.23 | 16664.69 |
4 cores and 8 GB | 64 | 800000 | 100 | 1377.89 | 27558.02 |
8 cores and 16 GB | 256 | 800000 | 300 | 2510.6 | 50212.25 |
8 cores and 32 GB | 256 | 800000 | 300 | 2767.62 | 55352.74 |
8 cores and 64 GB | 128 | 800000 | 450 | 2908.69 | 58174.12 |
16 cores and 64 GB | 256 | 800000 | 450 | 5379.72 | 107595.01 |
16 cores and 96 GB | 256 | 800000 | 600 | 5591.88 | 111838.18 |
16 cores and 128 GB | 256 | 800000 | 600 | 6304.74 | 126095.54 |
24 cores and 144 GB | 512 | 5000000 | 300 | 6935.08 | 138702.38 |
32 cores and 128 GB | 256 | 5000000 | 300 | 7742.32 | 154847.21 |
32 cores and 256 GB | 256 | 5000000 | 400 | 8341.2 | 166824.9 |
64 cores and 256 GB | 256 | 5000000 | 400 | 14683.84 | 293678.41 |
64 cores and 512 GB | 256 | 6000000 | 450 | 16363.74 | 327276.65 |
88 cores and 710 GB | 256 | 7000000 | 450 | 23149.55 | 462993.59 |
Scenario 4: Mixed Read-Write (point select)
Specifications | Concurrency | Data Volume of a Single Table (table_size) | Total Number of Tables (tables) | TPS | QPS |
2 cores and 4 GB | 64 | 800000 | 100 | 1350.51 | 21608.23 |
4 cores and 8 GB | 128 | 800000 | 100 | 2578.75 | 41260.1 |
8 cores and 16 GB | 256 | 800000 | 300 | 4567.45 | 73079.37 |
8 cores and 32 GB | 256 | 800000 | 300 | 5176.68 | 82827.15 |
8 cores and 64 GB | 128 | 800000 | 450 | 5436.25 | 86980.3 |
16 cores and 64 GB | 256 | 800000 | 450 | 9754.57 | 156073.6 |
16 cores and 96 GB | 256 | 800000 | 600 | 10262.53 | 164201.05 |
16 cores and 128 GB | 256 | 800000 | 600 | 11744.35 | 187910.13 |
24 cores and 144 GB | 256 | 5000000 | 300 | 12239.25 | 194828 |
32 cores and 128 GB | 256 | 5000000 | 300 | 13665.11 | 218642.4 |
32 cores and 256 GB | 256 | 5000000 | 400 | 14439.52 | 231033.03 |
64 cores and 256 GB | 512 | 5000000 | 400 | 25610.36 | 409767.14 |
64 cores and 512 GB | 512 | 6000000 | 450 | 26243.25 | 419893.44 |
88 cores and 710 GB | 64 | 7000000 | 450 | 31985.65 | 511772.02 |
Scenario 5: Write-Only
Specifications | Concurrency | Data Volume of a Single Table (table_size) | Total Number of Tables (tables) | TPS | QPS |
2 cores and 4 GB | 128 | 800000 | 100 | 3381.33 | 20239.02 |
4 cores and 8 GB | 128 | 800000 | 100 | 7216.54 | 43194.73 |
8 cores and 16 GB | 128 | 800000 | 300 | 13005.36 | 77843.84 |
8 cores and 32 GB | 128 | 800000 | 300 | 13858.45 | 82950.02 |
8 cores and 64 GB | 256 | 800000 | 450 | 13893.04 | 83157.09 |
16 cores and 64 GB | 1024 | 800000 | 450 | 25077.31 | 150100.73 |
16 cores and 96 GB | 128 | 800000 | 600 | 25975.78 | 155478.54 |
16 cores and 128 GB | 1024 | 800000 | 600 | 26690.32 | 159755.43 |
24 cores and 144 GB | 1024 | 5000000 | 300 | 30023.68 | 179707.33 |
32 cores and 128 GB | 1024 | 5000000 | 300 | 37492 | 224409.11 |
32 cores and 256 GB | 1024 | 5000000 | 400 | 38533.23 | 230641.43 |
64 cores and 256 GB | 1024 | 5000000 | 400 | 53930.69 | 322803.26 |
64 cores and 512 GB | 1024 | 6000000 | 450 | 57587.52 | 344691.26 |
88 cores and 710 GB | 1500 | 7000000 | 450 | 76714.87 | 459178.43 |
Was this page helpful?
You can also Contact Sales or Submit a Ticket for help.
Yes
No
Feedback