本文介绍AnalyticDB for MySQL、MySQL、Presto、Spark以及Impala的性能测试结果。

测试样本 执行时间(s)
AnalyticDB for MySQL MySQL 执行时间提升倍数 Presto 执行时间提升倍数 Spark 执行时间提升倍数 Impala 执行时间提升/降低倍数
tpch_sql1 19.959 900.535 44.12 130.83 5.55 64.088 2.211 6.080 -0.695
tpch_sql2 0.889 4.135 3.65 44.94 49.55 41.219 45.366 18.660 19.990
tpch_sql3 4.368 156.144 34.75 121.21 26.75 41.913 8.595 93.240 20.346
tpch_sql4 7.506 34.811 3.64 115.68 14.41 24.764 2.299 138.240 17.417
tpch_sql5 4.563 55.333 11.13 - - 67.922 13.885 - -
tpch_sql6 0.186 166.924 896.44 91.62 491.58 10.462 55.247 1.440 6.742
tpch_sql7 3.444 81.284 22.60 143.34 40.62 162.78 46.265 136.350 38.591
tpch_sql8 6.701 92.881 12.86 - - 77.574 10.576 277.470 40.407
tpch_sql9 37.154 417.531 10.24 - - 100.33 1.700 300.480 7.087
tpch_sql10 4.503 322.436 70.60 127.17 27.24 46.066 9.230 44.670 8.920
tpch_sql11 1.056 21.790 19.63 39.91 36.79 49.354 45.737 2.380 1.254
tpch_sql12 1.270 14,656.916 11,539.88 120.28 93.71 21.242 15.726 5.260 3.142
tpch_sql13 7.303 517.782 69.90 34.43 3.71 34.51 3.725 88.200 11.077
tpch_sql14 0.634 43.885 68.22 94.69 148.35 20.322 31.054 7.980 11.587
tpch_sql15 1.922 149.901 76.99 179.16 92.22 23.899 11.434 3.650 0.899
tpch_sql16 1.940 45.487 22.45 25.92 12.36 79.557 40.009 7.180 2.701
tpch_sql17 5.128 20.985 3.09 210.68 40.08 116.033 21.627 46.650 8.097
tpch_sql18 26.591 236.179 7.88 - - 119.788 3.505 158.340 4.955
tpch_sql19 2.566 15.867 5.18 97.32 36.93 21.68 7.449 11.080 3.318
tpch_sql20 3.537 71.311 19.16 116.28 31.88 43.971 11.432 20.300 4.739
tpch_sql21 18.946 162.672 7.59 201.77 9.65 80.793 3.264 509.870 25.912
tpch_sql22 7.159 13.323 0.86 32.9 3.60 21.832 2.050 46.220 5.456
总时长(s) 167.325 18,188.112 107.70 2139.423 11.79 1270.099 6.591 1923.74 10.497
说明
  • Impala中不允许HAVING中有子查询,tpch_sql5没有执行。
  • Presto在部分示例中出现了OOM,无法完成查询。
更新图标1图表2