本文通过实际案例对比了使用bulk方式与rowbyrow方式进行数据表刷新的速度差异。结果显示，对于大型数据表，bulk方式比rowbyrow方式快约两倍，前者仅需15.76分钟，而后者则需36.99分钟。
没有特殊情况尽量用bulk，bulk方式相比row yb row方式会快很多。
Table 'testtable' in location 'test' (199241477 rows) and location 'test1' (198948686 rows) differed by 292791 inserts (compression=99.8%). This row-wise refresh took 36.99 minutes.
bulk方式:
Refreshed 'testtable' from location 'test' to location 'test2' (199241477 rows 228 wide compression=98.8%). This bulk refresh took 15.76 minutes.
以下是刷新同一张大表的时间对比：
row by row方式:Table 'testtable' in location 'test' (199241477 rows) and location 'test1' (198948686 rows) differed by 292791 inserts (compression=99.8%). This row-wise refresh took 36.99 minutes.
bulk方式:
Refreshed 'testtable' from location 'test' to location 'test2' (199241477 rows 228 wide compression=98.8%). This bulk refresh took 15.76 minutes.
扫一扫
6861
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
