Automating Disk Formatting with nfsFormatDisk in ScriptsFormatting disks automatically is a common task in data center provisioning, test environments, and large-scale deployments. When done correctly, automation reduces human error, speeds provisioning, and ensures consistent device states. This article explains how to automate disk formatting using the nfsFormatDisk utility (or similarly named tool) within scripts, covers safety considerations, shows examples for different environments, and offers best practices for production usage.
What nfsFormatDisk does (short overview)
nfsFormatDisk is a command-line utility used to prepare storage devices for use in NFS-related workflows. Typically it:
- Wipes or formats block devices.
- Creates filesystems of a specified type.
- Optionally labels or mounts the resulting filesystem.
- May include checks to avoid formatting active system disks.
Exact flags and behavior can vary by implementation or vendor; consult your local nfsFormatDisk manual or –help output for precise options before scripting.
Safety first: precautions before automating formatting
Formatting is destructive. Automating it increases the risk of data loss if a wrong device is targeted. Before integrating nfsFormatDisk into scripts, implement these safeguards:
- Require explicit device identifiers (e.g., /dev/sdb) rather than patterns like /dev/sd*.
- Use a dry-run or –no-act option if available.
- Verify device is unmounted and not in use (lsof, fuser, /proc/mounts).
- Confirm device serial/WWN or UUID matches expected hardware.
- Add interactive confirmation for non-forced runs or provide a clear –force flag only for CI/CD contexts.
- Keep backups and ensure critical data is replicated elsewhere.
- Log every operation with timestamps, invoking user, and command output.
Recommended script structure
A robust script around nfsFormatDisk should:
- Parse and validate input arguments.
- Confirm the targeted device is correct and safe to operate on.
- Run the format operation, capturing stdout/stderr and exit codes.
- Verify the filesystem after formatting and optionally mount/label.
- Report success/failure and write an audit log.
Example POSIX-compatible structure (conceptual steps — see concrete examples below):
- header: shebang, set -euo pipefail, usage function
- parse args: device, fs-type, label, force/dry-run
- pre-checks: existence, mounted check, size threshold, confirm WWN/UUID if required
- run nfsFormatDisk with chosen flags
- post-check: fsck or blkid to confirm filesystem, mount if requested
- logging: append to /var/log/nfsformat.log
Example: Bash script for one-off automation
This example demonstrates a safe, interactive bash script that wraps nfsFormatDisk. Adjust flags to match your nfsFormatDisk implementation.
#!/usr/bin/env bash set -euo pipefail DEVICE="${1:-}" FSTYPE="${2:-ext4}" LABEL="${3:-}" DRY_RUN=0 LOG="/var/log/nfsformat.log" timestamp(){ date --iso-8601=seconds; } usage(){ echo "Usage: $0 /dev/sdX [fstype=ext4] [label] --dry-run" exit 2 } if [[ -z "$DEVICE" ]]; then usage; fi if [[ "$DEVICE" == "--dry-run" || "$3" == "--dry-run" ]]; then DRY_RUN=1; fi if [[ ! -b "$DEVICE" ]]; then echo "Error: $DEVICE not a block device"; exit 1; fi # Prevent formatting root or boot devices by pattern case "$DEVICE" in /dev/sd[a-z]1|/dev/sda|/dev/nvme0n1) echo "Refusing to format likely system device: $DEVICE"; exit 1 ;; esac if mount | grep -q "^$DEVICE"; then echo "Error: $DEVICE is mounted. Unmount first."; exit 1 fi echo "$(timestamp) INFO: Preparing to format $DEVICE as $FSTYPE label='$LABEL' dry_run=$DRY_RUN" >> "$LOG" if [[ "$DRY_RUN" -eq 1 ]]; then echo "Dry run: nfsFormatDisk --device $DEVICE --fstype $FSTYPE --label '$LABEL'" exit 0 fi # Optional device identity check (requires udevadm/blkid) WWN_EXPECTED="" # fill in if you have a specific WWN if [[ -n "$WWN_EXPECTED" ]]; then WWN_ACTUAL=$(blkid -o value -s UUID "$DEVICE" || true) if [[ "$WWN_ACTUAL" != "$WWN_EXPECTED" ]]; then echo "Device identity mismatch: expected $WWN_EXPECTED got $WWN_ACTUAL"; exit 1 fi fi # Run format if nfsFormatDisk --device "$DEVICE" --fstype "$FSTYPE" --label "$LABEL"; then echo "$(timestamp) INFO: Formatting succeeded for $DEVICE" >> "$LOG" else echo "$(timestamp) ERROR: Formatting failed for $DEVICE" >> "$LOG" exit 1 fi # Verify filesystem sleep 1 if blkid "$DEVICE" | grep -q "$FSTYPE"; then echo "$(timestamp) INFO: Verification OK for $DEVICE" >> "$LOG" else echo "$(timestamp) ERROR: Verification failed for $DEVICE" >> "$LOG" exit 1 fi # Optionally mount to /mnt/<label> if [[ -n "$LABEL" ]]; then MOUNTPOINT="/mnt/$LABEL" mkdir -p "$MOUNTPOINT" mount "$DEVICE" "$MOUNTPOINT" echo "$(timestamp) INFO: Mounted $DEVICE at $MOUNTPOINT" >> "$LOG" fi echo "Done." Example: Non-interactive automation in CI/CD
In fully automated environments (e.g., cloud-init, Terraform local-exec, or CI runners), ensure:
- Use explicit –force flags only after thorough testing.
- Include cloud provider metadata checks (instance ID, attached volume ID) to avoid cross-instance mistakes.
- Run in an isolated ephemeral instance when possible.
Example (pseudo shell snippet):
# Pre-check: confirm attached volume ID via cloud metadata API ATTACHED_ID=$(curl -s http://169.254.169.254/latest/meta-data/block-device-mapping/ephemeral0) if [[ "$ATTACHED_ID" != "expected-id" ]]; then exit 1; fi nfsFormatDisk --device /dev/xvdb --fstype xfs --force --label data-volume Logging and auditability
- Centralize logs (syslog, journald, or remote logging endpoint).
- Capture stdout/stderr and exit codes from nfsFormatDisk.
- Include environment context: hostname, timestamp, operator, orchestration job ID, and any attached volume IDs.
- Rotate logs and keep them for the period your compliance rules demand.
Error handling and retries
- Treat non-zero exit codes as actionable failures.
- Retry transient failures (IO errors, temporary device busy) with exponential backoff and limited attempts.
- For repeatable failures, capture core dumps or device SMART data (smartctl) for diagnosis.
Pseudo-retry loop:
max=3; n=0 until nfsFormatDisk --device "$DEVICE" --fstype "$FSTYPE"; do n=$((n+1)) if [[ $n -ge $max ]]; then break; fi sleep $((2**n)) done Integration with orchestration tools
- Ansible: wrap the script in an Ansible module or use raw/shell with careful become and checks.
- Terraform: use null_resource + local-exec to trigger the script after volume attach.
- Kubernetes: use init containers to prepare attached block devices before mounting into pods (ensure node-level privileges and careful node selection).
Testing and validation
- Test in an isolated lab with identical device types and OS images.
- Use virtual disks (loop devices) for unit tests: losetup, mkfs, and teardown.
- Validate that formatted disks behave as expected under workload: fio/iozone, mount/unmount cycles, and stress tests.
- Add continuous tests to CI pipelines for any changes to formatting scripts.
Best practices summary
- Require explicit device identifiers.
- Use dry-run and identity checks before destructive operations.
- Log everything and centralize audit records.
- Use retries for transient errors but fail fast on obvious misconfiguration.
- Test in isolated environments and automate validation.
Automating disk formatting with nfsFormatDisk can save time and ensure consistency when done with careful safeguards. Start conservative (dry runs, identity checks), add automation only after repeated successful tests, and always treat formatting operations as high-risk actions that require strong logging and verification.
Leave a Reply