We have been monitoring some SQLServer: Memory Manager's metrics, and noticed that after DBCC CheckDB job, metric Database Cache Memory (KB) drops down significantly. If to be exact, it dropped from 140 GB cached DB memory to 60 GB. And after that, slowly ramp up again during the week. (Amount of "Free Memory KB", went from 20 to 100 GB right after CheckDB) DBCC CheckDB is run every Sunday, so Database Cache Memory has to ramp up back again each week What is the behavior of this ? Why CheckDB pushes database pages out of memory ? Second question is why "buffer cache hit ratio" did not change after DBCC CheckDB completes ? It was 99.99% on average and after DBCC CheckDB job it drops to ~98.00%, and returns back to 99% pretty fast while I expected "buffer cache hit ratio" to drop significantly because database data has to be read from storage to RAM again ?

I have SQL Server running on 32-core CPU, and I use Performance Monitor (perfmon.msc) to collect data on Processor: % Processor Time (Instance - _Total) and Process: % Processor Time (Instance - sqlservr.exe) So for example for day 2019/07/21 (Sunday), it shows average values: Processor: % Processor Time (Instance - _Total) - 3.959 Process: % Processor Time (Instance - sqlservr.exe) - 107.312 Why is such a difference ? I expected sqlservr.exe to show values like 2.900 or 3.100 - something like this, but not 107.312 How to correlate _Total to sqlservr.exe ? I am using MaxDOP limited to 8 server level setting on SQL Server, if this is any help