Updated on 2024-10-08 GMT+08:00

Performance Test Data

Write Performance Testing

Table 1 Data used for testing write performance of cluster instances

Instance Specifications

Concurrent Requests

Write Performance (unit: rows/sec)

4 vCPUs | 16 GB

20

123648.75

8 vCPUs | 32 GB

40

221034.80

16 vCPUs | 64 GB

80

348762.25

32 vCPUs | 128 GB

160

496511.06

Table 2 Data used for testing write performance of single-node instances

Instance Specifications

Concurrent Requests

Write Performance (unit: rows/sec)

4 vCPUs | 8 GB

10

50113

8 vCPUs | 16 GB

10

108781

16 vCPUs | 32 GB

20

158744

Query Performance Testing

The test data depends on service models and instance specifications.

Table 1 describes the service models involved in this test.

  • The test instance of the cluster type, with specifications of 4 vCPUs and 16 GB and concurrent requests of 20. All metrics in the following table are average values calculated from 1 million executions.
    Table 3 Test data

    Test Model

    OPS (unit: queries/sec)

    Average Latency (unit: ms)

    Minimum Latency (unit: ms)

    Maximum Latency (unit: ms)

    high-cpu-all

    710.81

    28.12

    9.35

    714.11

    single-groupby-1-8-1

    1308.56

    13.74

    2.56

    148.96

    single-groupby-1-1-1

    6393.67

    3.10

    1.43

    45.02

    cpu-max-all-1

    850.51

    23.49

    6.16

    715.23

  • The test instance of the cluster type, with specifications of 8 vCPUs and 32 GB and concurrent requests of 40. All metrics in the following table are average values calculated from 1 million executions.
    Table 4 Test data

    Test Model

    OPS (unit: queries/sec)

    Average Latency (unit: ms)

    Minimum Latency (unit: ms)

    Maximum Latency (unit: ms)

    high-cpu-all

    1236.46

    32.34

    9.70

    412.86

    single-groupby-1-8-1

    2663.19

    12.47

    2.58

    222.84

    single-groupby-1-1-1

    9696.13

    4.03

    1.56

    141.06

    cpu-max-all-1

    1406.48

    28.42

    8.97

    444.16

  • The test instance of the cluster type, with specifications of 16 vCPUs and 64 GB and concurrent requests of 80. All metrics in the following table are average values calculated from 1 million executions.
    Table 5 Test data

    Test Model

    OPS (unit: queries/sec)

    Average Latency (unit: ms)

    Minimum Latency (unit: ms)

    Maximum Latency (unit: ms)

    high-cpu-all

    2107.83

    37.94

    11.33

    252.74

    single-groupby-1-8-1

    4707.25

    15.40

    3.29

    225.18

    single-groupby-1-1-1

    17658.59

    4.44

    1.80

    51.16

    cpu-max-all-1

    2262.40

    35.35

    12.80

    247.85

  • The test instance of the cluster type, with specifications of 32 vCPUs and 128 GB and concurrent requests of 160. All metrics in the following table are average values calculated from 1 million executions.
    Table 6 Test data

    Test Model

    OPS (unit: queries/sec)

    Average Latency (unit: ms)

    Minimum Latency (unit: ms)

    Maximum Latency (unit: ms)

    high-cpu-all

    3468.89

    46.10

    19.14

    290.61

    single-groupby-1-8-1

    5107.15

    13.84

    3.58

    118.97

    single-groupby-1-1-1

    23023.11

    6.72

    1.80

    74.45

    cpu-max-all-1

    3715.62

    43.04

    14.24

    186.80

  • The test instance of the single-node type, with specifications of 4 vCPUs and 8 GB and concurrent requests of 10. All metrics in the following table are average values calculated from 1 million executions.
    Table 7 Test data

    Test Model

    OPS (unit: queries/sec)

    Average Latency (unit: ms)

    Minimum Latency (unit: ms)

    Maximum Latency (unit: ms)

    high-cpu-all

    423.07

    14.17

    5.24

    693.53

    single-groupby-1-8-1

    1278.77

    4.68

    2.01

    822.53

    single-groupby-1-1-1

    3138.4

    1.9

    1.1

    424.77

    cpu-max-all-1

    357.93

    16.75

    8.51

    992.06

  • The test instance of the single-node type, with specifications of 8 vCPUs and 16 GB and concurrent requests of 20. All metrics in the following table are average values calculated from 1 million executions.
    Table 8 Test data

    Test Model

    OPS (unit: queries/sec)

    Average Latency (unit: ms)

    Minimum Latency (unit: ms)

    Maximum Latency (unit: ms)

    high-cpu-all

    1031.77

    15.49

    5.37

    614.3

    single-groupby-1-8-1

    3082.18

    5.18

    2.12

    154.53

    single-groupby-1-1-1

    7604.41

    2.1

    0.96

    31.93

    cpu-max-all-1

    856.75

    18.66

    7.76

    573.18

  • The test instance of the single-node type, with specifications of 16 vCPUs and 32 GB and concurrent requests of 20. All metrics in the following table are average values calculated from 1 million executions.
    Table 9 Test data

    Test Model

    OPS (unit: queries/sec)

    Average Latency (unit: ms)

    Minimum Latency (unit: ms)

    Maximum Latency (unit: ms)

    high-cpu-all

    1646.46

    12.13

    4.4

    409.82

    single-groupby-1-8-1

    3909.19

    5.11

    2.15

    122.95

    single-groupby-1-1-1

    10340.02

    1.93

    1.02

    146.8

    cpu-max-all-1

    1181.3

    16.92

    7.79

    175.29