Fuck me forgot to change REDO, causing copying users to be super slow

. A user's test environment, thinking about duplicating users for functional and single-point performance testing. However, the amount of user data is large, as shown in the figure

2. It takes about 2 hours and 20 minutes to copy a user in the test environment.

Test the configuration of the virtual machine:

At first I noticed that awr has some row locks

I thought it was an index problem and also killed the pk index at the beginning

Also noticed that write and read basically work hard on the index.. It is very slow to process

3. As a result, I didn't notice the wait time in the front of the awr (thanks to Manager Fang, I watched it for a long time and didn't notice it)

My redo is only 50m WTF

Then I counted: I found that there must be a logswitch in less than 3 minutes... Then I switched at least 50 times in 2 hours and 20 minutes..

 

Empiricism seriously thinks that I have adjusted the redo of the server, but I have not adjusted it..

Mark this is too low..

 

Guess you like

Origin http://43.154.161.224:23101/article/api/json?id=325236591&siteId=291194637