Hi forks,
We have several servers with persistence enabled, and two of them have
small memory, so they are not enough to store the amount of data
allocated to them.
Now the problem is that SQL takes a long time to execute. We monitored
the disk IO of these two servers and used the following commands:
iostat -dxm 3
The results are as follows:
We see that the r/s value is very high, but the rMB/s is not very high.
We know that there is data exchange between memory and disk, but is this
normal?
Are there any parameters (including os) that can be optimized?