Oracle Database Discussions

Hi,

We are experiencing log file sync as our top wait. We aren't performing excessive commits so we are assuming it is an I/O issue.
Our provider argues it is our bad code that is causing this issue. With increased load we are seeing the redo log write times going up
over 400+ ms. So as a test we ran a table copy

create table <name> as (select * from <my_table>)

where <my_table> has 2 million rows. The thing is memory and CPU are both generally always under 10% utilization even
during spikes. They are arguing that the create table statement is performing a full table scan and that can also
degrade performance as well and bring the database to a crawl. But wouldnt we be seeing memory and CPU go
way up if this was true?

Thanks!