Be aware that after upgrading your compute resources running on CentOS 6 to CentOS 7, the DM-Cache might not be working.
Note: this doesn't affect new compute resources running on CentOS 7, only the existing one.
Perform the following steps to fix the issue:
1. Wipe the disk that is assigned to cache:
[email@example.com ~]# dd if=/dev/zero bs=1M oflag=direct count=16 of=/dev/sda status=progress 16+0 records in 16+0 records out 16777216 bytes (17 MB) copied, 0.182174 s, 92.1 MB/s
2. Restart groupmon or reboot the compute resource.
[firstname.lastname@example.org ~]# /etc/init.d/groupmon restart