Add manual Atlas USB backup and reminder services

This commit is contained in:
Fabio Scotto di Santolo
2026-09-24 08:55:52 +02:00
parent d4e40d423a
commit 361ee77d72
11 changed files with 548 additions and 1 deletions

View File

@@ -65,6 +65,8 @@ Ansible-driven personal infrastructure repo for Fedora and Void desktops, Fedora
`ansible-playbook ansible/site.yml --limit atlas --tags packages,borg --check --diff` `ansible-playbook ansible/site.yml --limit atlas --tags packages,borg --check --diff`
- Atlas Borg progress logging only: - Atlas Borg progress logging only:
`ansible-playbook ansible/site.yml --limit atlas --tags borg_logging --check --diff` `ansible-playbook ansible/site.yml --limit atlas --tags borg_logging --check --diff`
- Atlas manual offline USB backup and 45Drives Alerts reminder:
`ansible-playbook ansible/site.yml --limit atlas --tags usb_backup,usb_reminder --check --diff`
- Prometheus/Aegis WireGuard gateway: - Prometheus/Aegis WireGuard gateway:
`ansible-playbook ansible/site.yml --limit prometheus,aegis --tags wireguard --check --diff` `ansible-playbook ansible/site.yml --limit prometheus,aegis --tags wireguard --check --diff`
- DuckDNS config only: `ansible-playbook ansible/site.yml --limit prometheus --tags duckdns --check --diff` - DuckDNS config only: `ansible-playbook ansible/site.yml --limit prometheus --tags duckdns --check --diff`
@@ -194,7 +196,14 @@ scheduled retention prune and monthly scrub remain runtime checks.
- [ ] Run and evaluate Borg against the populated pool: duration, repository capacity, deduplication, and - [ ] Run and evaluate Borg against the populated pool: duration, repository capacity, deduplication, and
a subsequent incremental archive must be observed before relying on the offline USB test. a subsequent incremental archive must be observed before relying on the offline USB test.
- [ ] Add the UUID-bound offline USB backup with versioned rsync, locking, capacity checks, verification, - [ ] Add the UUID-bound offline USB backup with versioned rsync, locking, capacity checks, verification,
safe unmounting and a tested restore procedure; never trigger it for an arbitrary USB disk. safe unmounting and a tested restore procedure; never trigger it for an arbitrary USB disk. The
LUKS/ext4 identities were read-only verified; the manual service and 45Drives Alerts reminder timer were
deployed on Atlas. Interactive LUKS unlock is part of the manual service; only the reminder is
scheduled for the first Saturday of each month at 10:00 Europe/Rome via the existing 45Drives
notifier. A manual test produced an Alerts notification, not an email. The first USB attempt failed
on a `security.selinux` xattr and was interrupted; the xattr filter is deployed and the temporary
recursive snapshot, open LUKS mapper, and failed service state were cleaned up. No complete backup or
tested USB restore exists; do not mark this item complete yet.
- [ ] Test restores independently from a ZFS snapshot, Borg, and the offline USB backup before relying on - [ ] Test restores independently from a ZFS snapshot, Borg, and the offline USB backup before relying on
any backup path. any backup path.
- [ ] Add monitoring and alerting for pool health, scrub/resilver, SMART data, temperatures, free space, - [ ] Add monitoring and alerting for pool health, scrub/resilver, SMART data, temperatures, free space,

View File

@@ -378,6 +378,55 @@ Atlas runtime activation is complete: the initial backup and repository check su
to a temporary directory was validated against the live `Archive` tree, the recovery-key export was copied to a temporary directory was validated against the live `Archive` tree, the recovery-key export was copied
to offline storage, and the temporary snapshot and bind mounts were cleaned up. to offline storage, and the temporary snapshot and bind mounts were cleaned up.
The offline USB backup is deployed as a manual-only service (`atlas_manage_usb_backup: true`):
Ansible never formats, unlocks, mounts, backs up to, or schedules the disk. Atlas' existing USB disk was verified
read-only on 2026-09-23 as LUKS UUID `577b3c43-ea37-4611-81a9-39d555cdfbd4`, containing ext4 UUID
`758e2d2e-a427-4797-aad9-39c3a9f17c7e` through mapper `zpool-backup`. It was mounted at
`/mnt/zpool-backup` at inspection time. The service deliberately requires the verified mapper to be
**not mounted** before starting. When necessary, `systemd-ask-password` requests the LUKS passphrase
through the `systemctl start` password agent; it is piped directly to `cryptsetup` without saving it,
passing it as a command argument, or caching it. The service then mounts the disk privately, takes a recursive ZFS snapshot,
copies every dataset to a versioned `atlas/snapshots/<timestamp>/` directory using `rsync --link-dest`,
verifies the result with a checksum-based dry run, atomically updates `atlas/latest`, unmounts and closes
LUKS. A failed run never replaces `latest` or removes an earlier complete version. Borg and the USB
backup may run concurrently from separate snapshots; both reading the same pool can reduce throughput.
The USB copy preserves ACLs and extended attributes except `security.selinux`, which the target
SELinux policy must recreate during a restore; do not restore data into service paths without relabeling.
Old USB versions are not pruned automatically, to avoid deleting the only offline
copy without an explicitly chosen retention policy; capacity checks include an estimated transfer size
and a 10 GiB free-space reserve. The disk must be physically disconnected after a successful backup
to make the copy offline.
To check the USB backup and reminder configuration without starting a backup, run:
```bash
ANSIBLE_LOCAL_TEMP=/tmp/ansible-local \
ansible-playbook ansible/site.yml --limit atlas --tags usb_backup,usb_reminder --check --diff
```
Before the first **manual** service start, safely unmount the currently mounted
`/mnt/zpool-backup`; never run it on an arbitrary mounted disk. Future starts
can begin with the mapper closed: `sudo systemctl start atlas-usb-backup.service` prompts for the
passphrase interactively and then performs the backup. Neither the LUKS password nor a key file belongs
in Ansible. Inspect the run with
`sudo journalctl -fu atlas-usb-backup.service`. There is intentionally no timer. Independently test a
read-only mount and restore from `atlas/latest` into an empty temporary directory before marking the
USB recovery path complete. Only `atlas-usb-reminder.timer` is enabled, for the first Saturday of each
month at 10:00 Europe/Rome. Its warning notification uses the existing 45Drives Houston notifier.
A manual test confirmed a notification in 45Drives Alerts, **not** an email. The reminder service log
reports notification submission, not email delivery; the role does not depend on SMTP/OAuth settings.
The reminder never starts the backup. Check its schedule with
`systemctl list-timers atlas-usb-reminder.timer` and the result in 45Drives Alerts.
The timer was verified active with its first scheduled run at 2026-10-03 10:00 CEST. The USB backup
service was verified inactive after deployment; no successful backup or email delivery is claimed.
The first manual USB attempt on 2026-09-23 did not complete: rsync was denied while removing
`security.selinux` on the USB filesystem, then the interrupted service left its recursive
`atlas-usb-20260923T185748Z-2469168` snapshot and the `zpool-backup` LUKS mapper open. The
rsync xattr filter was deployed afterward. The incomplete USB directory was absent on inspection;
the exact failed snapshot was removed, the verified and unmounted mapper closed, and the service
failed state cleared. A final check found no remnant snapshot, mount, mapper, or staging directory.
The failed attempt is not a valid backup, and no USB restore has been tested.
A temporary Nextcloud deployment on Atlas is also planned before Uranus: it requires separately A temporary Nextcloud deployment on Atlas is also planned before Uranus: it requires separately
declared persistent application, database, and cache storage, Vault-backed credentials, NPM-only declared persistent application, database, and cache storage, Vault-backed credentials, NPM-only
publishing through Aegis, and defined backup, upgrade, and eventual migration procedures. Do not deploy publishing through Aegis, and defined backup, upgrade, and eventual migration procedures. Do not deploy

View File

@@ -83,6 +83,13 @@ atlas_borg_randomized_delay: 30m
atlas_borg_keep_daily: 30 atlas_borg_keep_daily: 30
atlas_borg_keep_weekly: 8 atlas_borg_keep_weekly: 8
atlas_borg_keep_monthly: 12 atlas_borg_keep_monthly: 12
atlas_manage_usb_backup: true
# Read-only lsblk verification on Atlas, 2026-09-23. Never store the LUKS password here.
atlas_usb_backup_luks_uuid: 577b3c43-ea37-4611-81a9-39d555cdfbd4
atlas_usb_backup_fs_uuid: 758e2d2e-a427-4797-aad9-39c3a9f17c7e
atlas_usb_backup_mapper_name: zpool-backup
atlas_manage_usb_reminder: true
atlas_usb_reminder_calendar: "Sat *-*-01..07 10:00:00 Europe/Rome"
atlas_manage_sharing: true atlas_manage_sharing: true
atlas_manage_media_stack: false atlas_manage_media_stack: false

View File

@@ -97,6 +97,17 @@ atlas_borg_cache_dir: /var/cache/atlas-borg
atlas_borg_lock_path: /var/lib/atlas-borg/backup.lock atlas_borg_lock_path: /var/lib/atlas-borg/backup.lock
atlas_borg_recovery_export_path: "{{ playbook_dir }}/../secrets/recovery/atlas-borg-repokey.export" atlas_borg_recovery_export_path: "{{ playbook_dir }}/../secrets/recovery/atlas-borg-repokey.export"
# Manual-only offline backup. No USB device is formatted or mounted by Ansible.
atlas_manage_usb_backup: false
atlas_usb_backup_luks_uuid: ""
atlas_usb_backup_fs_uuid: ""
atlas_usb_backup_mapper_name: atlas-usb-backup
atlas_usb_backup_min_free_bytes: 10737418240
atlas_usb_backup_snapshot_prefix: atlas-usb
atlas_manage_usb_reminder: false
atlas_usb_reminder_calendar: ""
atlas_usb_reminder_notifier: /opt/45drives/houston/houston-notify
atlas_archive_mountpoint: "{{ atlas_mount_root }}/{{ atlas_zfs_dataset_archive }}" atlas_archive_mountpoint: "{{ atlas_mount_root }}/{{ atlas_zfs_dataset_archive }}"
atlas_services_mountpoint: "{{ atlas_mount_root }}/{{ atlas_zfs_dataset_services }}" atlas_services_mountpoint: "{{ atlas_mount_root }}/{{ atlas_zfs_dataset_services }}"
atlas_app_data_mountpoint: "{{ atlas_mount_root }}/{{ atlas_zfs_dataset_app_data }}" atlas_app_data_mountpoint: "{{ atlas_mount_root }}/{{ atlas_zfs_dataset_app_data }}"

View File

@@ -20,6 +20,9 @@
- name: Import Atlas Borg backup tasks - name: Import Atlas Borg backup tasks
ansible.builtin.import_tasks: borg_backup.yml ansible.builtin.import_tasks: borg_backup.yml
- name: Import Atlas offline USB backup tasks
ansible.builtin.import_tasks: usb_backup.yml
- name: Import Atlas file sharing tasks - name: Import Atlas file sharing tasks
ansible.builtin.import_tasks: sharing.yml ansible.builtin.import_tasks: sharing.yml

View File

@@ -0,0 +1,135 @@
---
- name: Validate Atlas offline USB backup configuration
tags: [atlas, storage, backup, usb_backup]
ansible.builtin.assert:
that:
- atlas_manage_storage | bool
- atlas_zfs_pool != 'CHANGEME_ZFS_POOL'
- atlas_mount_root.startswith('/')
- atlas_usb_backup_luks_uuid is match('^[0-9a-fA-F]{8}(-[0-9a-fA-F]{4}){3}-[0-9a-fA-F]{12}$')
- atlas_usb_backup_fs_uuid is match('^[0-9a-fA-F]{8}(-[0-9a-fA-F]{4}){3}-[0-9a-fA-F]{12}$')
- atlas_usb_backup_luks_uuid != atlas_usb_backup_fs_uuid
- atlas_usb_backup_mapper_name is match('^[a-z][a-z0-9_-]*$')
- atlas_usb_backup_min_free_bytes | int > 0
- atlas_usb_backup_snapshot_prefix is match('^[a-z0-9][a-z0-9_-]*$')
- atlas_usb_backup_snapshot_prefix != atlas_borg_snapshot_prefix
- atlas_usb_backup_snapshot_prefix != atlas_zfs_snapshot_prefix
fail_msg: >-
The manual Atlas USB backup needs verified LUKS and ext4 UUIDs, a safe
mapper name, positive free-space reserve, and a unique snapshot prefix.
when: atlas_manage_usb_backup | bool
- name: Install rsync for the Atlas offline USB backup
tags: [atlas, storage, backup, usb_backup]
ansible.builtin.dnf:
name: rsync
state: present
when: atlas_manage_usb_backup | bool
- name: Install the manual Atlas offline USB backup helper
tags: [atlas, storage, backup, usb_backup]
ansible.builtin.template:
src: atlas-usb-backup.sh.j2
dest: /usr/local/sbin/atlas-usb-backup
owner: root
group: root
mode: "0750"
when: atlas_manage_usb_backup | bool
- name: Install the manual Atlas offline USB backup service
tags: [atlas, storage, backup, usb_backup]
ansible.builtin.template:
src: atlas-usb-backup.service.j2
dest: /etc/systemd/system/atlas-usb-backup.service
owner: root
group: root
mode: "0644"
when: atlas_manage_usb_backup | bool
- name: Reload systemd for the Atlas offline USB backup service
tags: [atlas, storage, backup, usb_backup]
ansible.builtin.systemd:
daemon_reload: true
when:
- atlas_manage_usb_backup | bool
- not ansible_check_mode
- name: Validate the 45Drives Atlas USB reminder configuration
tags: [atlas, backup, usb_reminder]
ansible.builtin.assert:
that:
- atlas_manage_usb_backup | bool
- atlas_usb_reminder_calendar | length > 0
- atlas_usb_reminder_notifier.startswith('/opt/45drives/houston/')
fail_msg: >-
Enable the manual USB backup and declare a systemd calendar before
enabling its 45Drives Alerts reminder.
when: atlas_manage_usb_reminder | bool
- name: Validate the Atlas USB reminder calendar
tags: [atlas, backup, usb_reminder]
ansible.builtin.command:
argv:
- systemd-analyze
- calendar
- "{{ atlas_usb_reminder_calendar }}"
changed_when: false
check_mode: false
when: atlas_manage_usb_reminder | bool
- name: Inspect the existing 45Drives notifier
tags: [atlas, backup, usb_reminder]
ansible.builtin.stat:
path: "{{ atlas_usb_reminder_notifier }}"
register: atlas_usb_reminder_notifier_file
when: atlas_manage_usb_reminder | bool
- name: Require the configured 45Drives notifier for USB reminders
tags: [atlas, backup, usb_reminder]
ansible.builtin.assert:
that:
- atlas_usb_reminder_notifier_file.stat.executable | default(false)
fail_msg: >-
The existing 45Drives Houston notifier must be executable.
when: atlas_manage_usb_reminder | bool
- name: Install the 45Drives Atlas USB reminder helper
tags: [atlas, backup, usb_reminder]
ansible.builtin.template:
src: atlas-usb-reminder.py.j2
dest: /usr/local/libexec/atlas-usb-reminder
owner: root
group: root
mode: "0750"
when: atlas_manage_usb_reminder | bool
- name: Install the 45Drives Atlas USB reminder service
tags: [atlas, backup, usb_reminder]
ansible.builtin.template:
src: atlas-usb-reminder.service.j2
dest: /etc/systemd/system/atlas-usb-reminder.service
owner: root
group: root
mode: "0644"
when: atlas_manage_usb_reminder | bool
- name: Install the 45Drives Atlas USB reminder timer
tags: [atlas, backup, usb_reminder]
ansible.builtin.template:
src: atlas-usb-reminder.timer.j2
dest: /etc/systemd/system/atlas-usb-reminder.timer
owner: root
group: root
mode: "0644"
when: atlas_manage_usb_reminder | bool
- name: Enable only the Atlas USB notification reminder timer
tags: [atlas, backup, usb_reminder]
ansible.builtin.systemd:
name: atlas-usb-reminder.timer
enabled: true
state: started
daemon_reload: true
when:
- atlas_manage_usb_reminder | bool
- not ansible_check_mode

View File

@@ -0,0 +1,29 @@
[Unit]
Description=Run a manual, UUID-bound offline USB backup of Atlas ZFS datasets
Requires=zfs.target
After=zfs.target
ConditionFileIsExecutable=/usr/local/sbin/atlas-usb-backup
[Service]
Type=oneshot
ExecStart=/usr/local/sbin/atlas-usb-backup
User=root
Group=root
UMask=0077
TimeoutStartSec=infinity
RuntimeDirectory=atlas-usb-backup
RuntimeDirectoryMode=0700
Nice=15
IOSchedulingClass=best-effort
IOSchedulingPriority=7
PrivateMounts=true
PrivateTmp=true
ProtectHome=true
ProtectSystem=strict
ReadWritePaths=/run/atlas-usb-backup /run/lock
ProtectKernelTunables=true
ProtectKernelModules=true
ProtectControlGroups=true
RestrictAddressFamilies=AF_UNIX
RestrictRealtime=true
LockPersonality=true

View File

@@ -0,0 +1,241 @@
#!/usr/bin/env bash
set -Eeuo pipefail
export LC_ALL=C.utf8
export PATH=/usr/sbin:/usr/bin:/sbin:/bin
readonly pool={{ atlas_zfs_pool | quote }}
readonly mount_root={{ atlas_mount_root | quote }}
readonly luks_uuid={{ atlas_usb_backup_luks_uuid | quote }}
readonly fs_uuid={{ atlas_usb_backup_fs_uuid | quote }}
readonly mapper_name={{ atlas_usb_backup_mapper_name | quote }}
readonly snapshot_prefix={{ atlas_usb_backup_snapshot_prefix | quote }}
readonly min_free_bytes={{ atlas_usb_backup_min_free_bytes | int }}
readonly mapper="/dev/mapper/${mapper_name}"
readonly outer="/dev/disk/by-uuid/${luks_uuid}"
readonly runtime_dir=/run/atlas-usb-backup
readonly source_dir="${runtime_dir}/source"
readonly usb_mount="${runtime_dir}/target"
readonly backup_root="${usb_mount}/atlas"
snapshot_name=""
snapshot_created=false
mapper_opened_by_script=false
usb_mounted=false
published=false
partial=""
mounted_targets=()
# shellcheck disable=SC2329
cleanup() {
local status=$?
local cleanup_status=0
local index
local source_mount_failed=false
trap - EXIT HUP INT TERM
set +e
if [[ -n "$partial" && "$published" == false && "$usb_mounted" == true ]]; then
rm -rf -- "$partial" || cleanup_status=2
fi
if [[ "$usb_mounted" == true ]]; then
umount "$usb_mount" || cleanup_status=2
fi
{% raw %}
for ((index = ${#mounted_targets[@]} - 1; index >= 0; index--)); do
{% endraw %}
if mountpoint -q "${mounted_targets[$index]}"; then
umount "${mounted_targets[$index]}" || source_mount_failed=true
fi
if ! mountpoint -q "${mounted_targets[$index]}"; then
rmdir -- "${mounted_targets[$index]}" 2>/dev/null || true
fi
done
if [[ "$source_mount_failed" == false ]]; then
rmdir -- "$source_dir" 2>/dev/null || true
else
cleanup_status=2
printf 'Source bind mount cleanup failed; keeping the snapshot for recovery\n' >&2
fi
if [[ "$snapshot_created" == true && "$source_mount_failed" == false ]]; then
flock 9
zfs destroy -r "${pool}@${snapshot_name}" || cleanup_status=2
flock -u 9
fi
if [[ "$usb_mounted" == true || "$mapper_opened_by_script" == true ]] &&
! mountpoint -q "$usb_mount" &&
! findmnt -rn -S "$mapper" >/dev/null; then
cryptsetup close "$mapper_name" || cleanup_status=2
fi
rmdir -- "$usb_mount" 2>/dev/null || true
if ((status == 0 && cleanup_status != 0)); then
status=$cleanup_status
fi
exit "$status"
}
trap cleanup EXIT
trap 'exit 143' HUP INT TERM
exec 8>/run/lock/atlas-usb-backup.lock
flock -n 8 || { printf 'Atlas USB backup is already running\n' >&2; exit 75; }
exec 9>/run/lock/atlas-zfs-snapshot.lock
zpool list -H -o name "$pool" >/dev/null
[[ -b "$outer" ]] || { printf 'Configured LUKS UUID is not connected\n' >&2; exit 66; }
[[ "$(blkid -s TYPE -o value "$outer")" == crypto_LUKS ]] || {
printf 'Configured outer UUID is not a LUKS container\n' >&2
exit 65
}
[[ "$(blkid -s UUID -o value "$outer")" == "$luks_uuid" ]] || exit 65
if ! cryptsetup status "$mapper_name" >/dev/null; then
printf 'Requesting the LUKS passphrase for the configured USB disk\n'
systemd-ask-password -n --no-tty --timeout=300 \
--id="atlas-usb-backup:${luks_uuid}" \
'Atlas offline USB backup LUKS passphrase:' |
cryptsetup open --type luks2 --key-file - "$outer" "$mapper_name"
mapper_opened_by_script=true
fi
backing_device="$(cryptsetup status "$mapper_name" | awk '$1 == "device:" { print $2 }')"
[[ -n "$backing_device" && "$(readlink -f "$backing_device")" == "$(readlink -f "$outer")" ]] || {
printf 'The unlocked mapper does not belong to the configured LUKS UUID\n' >&2
exit 65
}
[[ "$(blkid -s TYPE -o value "$mapper")" == ext4 ]] || {
printf 'The unlocked USB filesystem is not ext4\n' >&2
exit 65
}
[[ "$(blkid -s UUID -o value "$mapper")" == "$fs_uuid" ]] || {
printf 'The unlocked USB filesystem UUID does not match\n' >&2
exit 65
}
if findmnt -rn -S "$mapper" >/dev/null; then
printf 'The USB filesystem is already mounted elsewhere\n' >&2
exit 65
fi
[[ ! -e "$source_dir" && ! -e "$usb_mount" ]] || {
printf 'USB backup staging directories already exist; inspect them manually\n' >&2
exit 65
}
mkdir -m 0700 "$usb_mount"
mount -t ext4 -o nodev,nosuid,noexec "$mapper" "$usb_mount"
usb_mounted=true
[[ "$(readlink -f "$(findmnt -nro SOURCE --target "$usb_mount")")" == "$(readlink -f "$mapper")" ]] || {
printf 'Mounted USB source does not match the verified mapper\n' >&2
exit 65
}
for path in "$backup_root" "$backup_root/snapshots"; do
[[ ! -L "$path" ]] || { printf 'Unsafe symlink in USB backup destination\n' >&2; exit 65; }
mkdir -p -- "$path"
[[ -d "$path" ]] || exit 65
chown root:root -- "$path"
chmod 0700 -- "$path"
done
free_bytes="$(df -B1 --output=avail "$usb_mount" | tail -n 1 | tr -d ' ')"
if ((free_bytes < min_free_bytes)); then
printf 'USB free space (%s bytes) is below the required reserve (%s bytes)\n' \
"$free_bytes" "$min_free_bytes" >&2
exit 73
fi
mkdir -m 0700 "$source_dir"
flock 9
timestamp="$(date -u +%Y%m%dT%H%M%SZ)"
snapshot_name="${snapshot_prefix}-${timestamp}-$$"
zfs snapshot -r "${pool}@${snapshot_name}"
snapshot_created=true
flock -u 9
printf 'Created recursive USB source snapshot %s@%s\n' "$pool" "$snapshot_name"
while IFS=$'\t' read -r dataset dataset_mountpoint mounted; do
if [[ "$mounted" != yes ]]; then
printf 'Dataset %s is not mounted; refusing an incomplete backup\n' "$dataset" >&2
exit 65
fi
if [[ "$dataset_mountpoint" != "$mount_root" && "$dataset_mountpoint" != "$mount_root/"* ]]; then
printf 'Dataset %s has unexpected mountpoint %s\n' "$dataset" "$dataset_mountpoint" >&2
exit 65
fi
dataset_suffix="${dataset#"$pool"}"
source_path="${dataset_mountpoint}/.zfs/snapshot/${snapshot_name}"
target_path="${source_dir}${dataset_suffix}"
mkdir -p "$target_path"
mount --bind "$source_path" "$target_path"
mounted_targets+=("$target_path")
mount -o remount,bind,ro "$target_path"
done < <(zfs list -H -o name,mountpoint,mounted -s name -r "$pool")
previous=""
if [[ -e "$backup_root/latest" || -L "$backup_root/latest" ]]; then
[[ -L "$backup_root/latest" ]] || { printf 'latest is not a symlink\n' >&2; exit 65; }
previous="$(readlink -e "$backup_root/latest")"
[[ -n "$previous" && "$previous" == "$backup_root/snapshots/"* && -d "$previous" ]] || {
printf 'latest does not point to a complete snapshot on the USB disk\n' >&2
exit 65
}
fi
backup_name="${timestamp}-$$"
candidate_partial="${backup_root}/snapshots/.incomplete-${backup_name}"
complete="${backup_root}/snapshots/${backup_name}"
[[ ! -e "$candidate_partial" && ! -L "$candidate_partial" && ! -e "$complete" && ! -L "$complete" ]] || exit 65
mkdir -m 0700 "$candidate_partial"
partial="$candidate_partial"
printf 'Copying the consistent pool tree to USB backup %s\n' "$backup_name"
# Preserve ACLs and other xattrs, but let SELinux relabel restored data on the
# destination host instead of trying to write source security.selinux labels to USB.
rsync_args=(-aHAXS --numeric-ids '--filter=-x security.selinux' "--info=progress2,stats2")
estimate_args=(-aHAXS --numeric-ids '--filter=-x security.selinux' --dry-run --stats)
if [[ -n "$previous" ]]; then
rsync_args+=("--link-dest=$previous")
estimate_args+=("--link-dest=$previous")
fi
# The rsync dry run estimates changed file bytes after link-dest deduplication.
# Metadata and filesystem allocation still require the separate free-space reserve.
estimate="$(rsync "${estimate_args[@]}" "${source_dir}/" "${partial}/")"
transfer_bytes="$(printf '%s\n' "$estimate" | awk -F: \
'/^Total transferred file size:/ { gsub(/[^0-9]/, "", $2); print $2 }')"
[[ "$transfer_bytes" =~ ^[0-9]+$ ]] || {
printf 'Could not determine the USB transfer size\n' >&2
exit 74
}
if ((free_bytes - transfer_bytes < min_free_bytes)); then
printf 'Insufficient USB space: %s bytes free, %s estimated transfer, %s reserved\n' \
"$free_bytes" "$transfer_bytes" "$min_free_bytes" >&2
exit 73
fi
rsync "${rsync_args[@]}" "${source_dir}/" "${partial}/"
printf 'Verifying USB backup %s with a checksum-based dry run\n' "$backup_name"
verification="${runtime_dir}/verification.out"
rsync -aHAXS --numeric-ids --filter='-x security.selinux' \
--checksum --dry-run --delete --itemize-changes \
"${source_dir}/" "${partial}/" >"$verification"
if [[ -s "$verification" ]]; then
printf 'USB verification found mismatches; refusing to publish the backup\n' >&2
exit 74
fi
free_bytes="$(df -B1 --output=avail "$usb_mount" | tail -n 1 | tr -d ' ')"
if ((free_bytes < min_free_bytes)); then
printf 'USB backup completed below the free-space reserve; refusing to publish it\n' >&2
exit 73
fi
mv -- "$partial" "$complete"
partial=""
ln -s "snapshots/${backup_name}" "${backup_root}/.latest-${backup_name}"
mv -Tf -- "${backup_root}/.latest-${backup_name}" "${backup_root}/latest"
published=true
sync -f "$complete"
sync -f "$backup_root"
printf 'USB backup %s verified and published; unmounting and closing LUKS\n' "$backup_name"

View File

@@ -0,0 +1,31 @@
#!/usr/bin/python3
"""Submit a manual-backup reminder through Atlas' existing Houston notifier."""
import json
import subprocess
from datetime import datetime, timezone
now = datetime.now(timezone.utc)
message = {
"timestamp": now.isoformat(timespec="seconds"),
"unixtime": int(now.timestamp()),
"event": "atlas_usb_backup_reminder",
"severity": "warning",
"subject": "Promemoria backup USB offline Atlas",
"email_message": (
"Collega il disco USB di backup ad Atlas ed esegui manualmente il backup offline.\n"
"Il promemoria non avvia il backup. Controlla che il disco non sia\n"
"montato; poi esegui:\n\n"
" sudo systemctl start atlas-usb-backup.service\n\n"
"Verifica l'esito con:\n"
" sudo journalctl -u atlas-usb-backup.service -n 100 --no-pager\n\n"
"Dopo la riuscita, scollega fisicamente il disco."
),
}
subprocess.run(
[{{ atlas_usb_reminder_notifier | to_json }}, json.dumps(message)],
check=True,
)
print("Atlas USB backup reminder submitted to 45Drives Alerts; email delivery is not verified.", flush=True)

View File

@@ -0,0 +1,22 @@
[Unit]
Description=45Drives Alerts reminder to run the manual Atlas offline USB backup
Requires=houston-dbus.service
After=houston-dbus.service
ConditionFileIsExecutable=/usr/local/libexec/atlas-usb-reminder
[Service]
Type=oneshot
ExecStart=/usr/local/libexec/atlas-usb-reminder
User=root
Group=root
UMask=0077
NoNewPrivileges=true
PrivateTmp=true
ProtectHome=true
ProtectSystem=strict
ProtectKernelTunables=true
ProtectKernelModules=true
ProtectControlGroups=true
RestrictAddressFamilies=AF_UNIX
RestrictRealtime=true
LockPersonality=true

View File

@@ -0,0 +1,10 @@
[Unit]
Description=Remind the administrator to run the manual Atlas offline USB backup
[Timer]
OnCalendar={{ atlas_usb_reminder_calendar }}
Persistent=true
Unit=atlas-usb-reminder.service
[Install]
WantedBy=timers.target