[Method Improvement] Update `DiskSetup.sh` to Ensure That `parted` or `wipefs` Are Consistently Used for Clearing Pattition Tables

[ferricoxide]

Describe the bug

With RHEL 9, in particular, use of the:

dd if=/dev/zero of="${CHROOTDEV}" bs=512 count=1000 > /dev/null 2>&1 || \

Is contraindicated. Better methods are:

parted -sf "${CHROOTDEV}" rm "${PART_NUM}"

Or:

wipefs -a "${CHROOTDEV}" 

The parted -sf ... method is used in the script, but not universally.

Severity

• Moderately Broken (Trivial work-around)

To Reproduce
Steps to reproduce the behavior:

1. Run AMIgen9
2. Observe script-outputs
3. Note that outputs include a block like:

   [...ELIDED...]
   logger -i -t DiskSetup.sh -p kern.crit -s -- 'Clearing existing partition-tables...'
   [[ NONE =~ ^[0-9]+$ ]]
   dd if=/dev/zero of=/dev/sda bs=512 count=1000
   partprobe /dev/sda
   [...ELIDED...]

   (Example from Azure)
4. Build-script may or may not error-exit after this output

Expected behavior

All partitions should be removed using either parted -s <DEVICE> rm <TARGET_SLICE> or, per updated guidance from Red Hat, wipefs -a <DEVICE> .

Deviance Description

Previous efforts to remove/replace use of the dd-based method seem to be incomplete.

Screenshots

Additional context

Fix Suggestions

Update all actions attempting to clear the target-device's partitions to us the parted or wipefs method.