tencent cloud

Feedback

Test Results - Full Cache Scenario

Last updated: 2022-04-13 12:04:10

    This document lists the performance comparison test results between TDSQL-C for MySQL and TencentDB for MySQL in the full cache scenario.

    Full Cache Scenario Overview

    In the full cache scenario, as all data can be put into the cache, the disk doesn't need to be read and written to update the cache during queries.

    Full Cache Scenario Test Conclusion

    • The higher the instance specification, the more obvious the performance advantage of TDSQL-C for MySQL. TencentDB for MySQL's write and read-write performance reaches a bottleneck at the 32-core specification, but TDSQL-C for MySQL can further increase the QPS with more CPU cores.
    • In most scenarios, TDSQL-C for MySQL can achieve a CPU utilization of above 90% on compute nodes. The test shows that the resource utilization of TDSQL-C for MySQL is better than that of TencentDB for MySQL.
    • TDSQL-C for MySQL is more stable in terms of request delay (RTT) and almost doesn't jitter at all when the dataset is fully cached.
    Dataset CharacteristicsTest ScenarioRead TypeConclusion
    Full cacheWrite-TDSQL-C for MySQL has a higher performance
    ReadPOINT SELECTTDSQL-C for MySQL has a higher performance
    ReadRANGE SELECTTDSQL-C for MySQL and TencentDB for MySQL have generally the same performance under low specifications, but the latter outperforms the former under high specifications
    Read-writePOINT SELECTTDSQL-C for MySQL has a higher performance
    Read-writeRANGE SELECTTDSQL-C for MySQL and TencentDB for MySQL have generally the same performance under most specifications

    Full Cache Scenario Test Results

    Scenario 1: Write

    Specification Concurrency Table Size Tables QPS
    TencentDB for MySQL TDSQL-C for MySQL
    2-core 16 GB MEM96250002501566530054
    4-core 16 GB MEM192250002504057453334
    4-core 32 GB MEM192250002504296653713
    8-core 32 GB MEM2562500025067229100737
    8-core 64 GB MEM256250002507695599480
    16-core 64 GB MEM51225000250134590181035
    16-core 96 GB MEM51225000250142419181029
    16-core 128 GB MEM51225000250144529181482
    32-core 128 GB MEM100025000250224786319913
    32-core 256 GB MEM100025000250220350370294
    64-core 256 GB MEM100025000250236079448221

    Scenario 2: Read (POINT SELECT)

    Specification Concurrency Table Size Tables QPS
    TencentDB for MySQL TDSQL-C for MySQL
    2-core 16 GB MEM1500250002503863357153
    4-core 16 GB MEM15002500025080398108368
    4-core 32 GB MEM15002500025081100108639
    8-core 32 GB MEM150025000250159885185710
    8-core 64 GB MEM150025000250172800206007
    16-core 64 GB MEM200025000250313223402101
    16-core 96 GB MEM200025000250321229402101
    16-core 128 GB MEM200025000250321617403809
    32-core 128 GB MEM200025000250409118715886
    32-core 256 GB MEM200025000250549297719295
    64-core 256 GB MEM2000250002506700261125180

    Scenario 3: Read (RANGE SELECT)

    Specification Concurrency Table Size Tables QPS
    TencentDB for MySQL TDSQL-C for MySQL
    2-core 16 GB MEM64250002501447415837
    4-core 16 GB MEM64250002503152331169
    4-core 32 GB MEM64250002503198831048
    8-core 32 GB MEM64250002505036259531
    8-core 64 GB MEM64250002506368659008
    16-core 64 GB MEM12825000250113098113779
    16-core 96 GB MEM12825000250124928113377
    16-core 128 GB MEM12825000250128728113606
    32-core 128 GB MEM25625000250212540197144
    32-core 256 GB MEM25625000250199970197796
    64-core 256 GB MEM25625000250304502289460

    Scenario 4: Read-write (POINT SELECT)

    Specification Concurrency Table Size Tables QPS
    TencentDB for MySQL TDSQL-C for MySQL
    2-core 16 GB MEM512250002502255438001
    4-core 16 GB MEM512250002505784171560
    4-core 32 GB MEM512250002506079772625
    8-core 32 GB MEM5122500025097243138499
    8-core 64 GB MEM51225000250112912138784
    16-core 64 GB MEM51225000250198182247586
    16-core 96 GB MEM51225000250203479247631
    16-core 128 GB MEM51225000250209947248301
    32-core 128 GB MEM51225000250348721400126
    32-core 256 GB MEM51225000250333932379531
    64-core 256 GB MEM51225000250439984553040

    Scenario 5: Read-write (RANGE SELECT)

    Specification Concurrency Table Size Tables QPS
    TencentDB for MySQL TDSQL-C for MySQL
    2-core 16 GB MEM64250002501356816072
    4-core 16 GB MEM256250002503331834553
    4-core 32 GB MEM256250002503376634321
    8-core 32 GB MEM256250002505558865190
    8-core 64 GB MEM256250002506261665701
    16-core 64 GB MEM25625000250111148123984
    16-core 96 GB MEM25625000250131182124659
    16-core 128 GB MEM38425000250130767125947
    32-core 128 GB MEM38425000250218580214861
    32-core 256 GB MEM38425000250210922216303
    64-core 256 GB MEM38425000250308399312941
    Contact Us

    Contact our sales team or business advisors to help your business.

    Technical Support

    Open a ticket if you're looking for further assistance. Our Ticket is 7x24 avaliable.

    7x24 Phone Support