CentOS / RHEL horrible disk performance with SSD

noop, is essentially a first-in first-out (FIFO) queue with no extra logic.

Each virtual machine can stop worrying about the disk, instead passing I/O requests along to the hypervisor to make a better decision about overall performance.

Add “elevator=noop” to the kernel parameters in your boot loader’s configuration in /etc/grub.conf

`echo noop > /sys/block/${DEVICE}/queue/scheduler`.

Set the default in your kernel config for all VM kernels you roll, with `CONFIG_DEFAULT_IOSCHED=”anticipatory”`.




VMware Virtual Machine Hosting