Also, the first snap were 10 minutes, this second one is 4 minutes!
However if you look the log file sync waits
The wait time of second SNAP with only around 13% of first SNAP waits is MUCH MUCH longer than the first SNAP, clearly shows you have serious I/O problems with your redo log I/O subsystemCode:First SNAP without CTAS: log file sync 56,757 225,427 63.60 Second SNAP with CTAS: log file sync 7,697 686,518 93.81
Code:SECOND SNAP Event Waits Timeouts Time (cs) (ms) /txn ---------------------------- ------------ ---------- ----------- ------ ------ log file sync 7,697 3,637 686,518 892 1.9 db file sequential read 87,472 0 28,547 3 21.4 log file parallel write 349 0 9,263 265 0.1 db file parallel write 256 0 2,547 99 0.1 log file switch completion 26 3 1,179 453 0.0 log buffer space 108 0 1,043 97 0.0 buffer busy waits 264 2 898 34 0.1 control file parallel write 82 0 677 83 0.0 direct path write 161 0 372 23 0.0 FIRST SNAP Event Waits Timeouts Time (cs) (ms) /txn ---------------------------- ------------ ---------- ----------- ------ ------ log file sync 56,757 0 225,427 40 1.0 db file sequential read 1,729,916 0 93,853 1 30.6 log file parallel write 39,154 0 29,790 8 0.7 db file parallel write 576 0 2,203 38 0.0 SQL*Net more data to client 540,735 0 2,161 0 9.6 control file parallel write 206 0 508 25 0.0 log file sequential read 4,194 0 163 0 0.1 log file switch completion 5 0 75 150 0.0




Reply With Quote