Description of problem: As a result of this package being introduced to Fedora it's found its way into RHEL via EPEL. Due to the udev rules that are bundled with this package, in addition to the following (https://github.com/amazonlinux/amazon-ec2-utils/issues/37), it can also have undesirable results especially if it's installed later in lifecycle of host. This scenario comes up when someone is just trying to get access to the cli tool `ebsnvme-id`, when something like `lsblk -f -t -m -o +SERIAL | sed 's/vol/vol-/'` or an equivalent alias would suffice. https://packages.fedoraproject.org/pkgs/amazon-ec2-utils/amazon-ec2-utils/epel-8.html https://rhel.pkgs.org/8/epel-x86_64/amazon-ec2-utils-2.2.0-1.el8.noarch.rpm.html https://github.com/amazonlinux/amazon-ec2-utils Version-Release number of selected component (if applicable): amazon-ec2-utils-2.2.0-1.el8.noarch Steps to Reproduce: 1. Configure LVM setup 2. Install amazon-ec2-utils from EPEL on RHEL 8.x 3. lvmdiskcan shows the newly attached pv as /dev/sdX where as previously they were /dev/nvmeXnX and it's not apparent to end user the reason why this is happening. Actual results: # lvmdiskscan /dev/nvme0n1p2 [ <50.00 GiB] ... /dev/nvme5n1p1 [ 1.00 TiB] LVM physical volume ... /dev/sdy [ 1.00 TiB] Additional info: Perhaps as an alternative package `ebsnvme-id` separately for those that just want the CLI tool.