screentinker/scripts/backup.sh
ScreenTinker 3ac81a4206 feat(ops): nightly backup script with point-in-time content history
Adds scripts/backup.sh — atomic SQLite .backup + hard-linked point-in-time
content snapshots, daily (7) + monthly (12) retention, and an error log.
Env-configurable (SCREENTINKER_DIR/BACKUP_DIR/DB/UPLOADS/*_KEEP*) so any
self-hoster can use it; defaults target a /opt/screentinker install.

Hardens two real failure modes found in production:
- Content snapshots EXCLUDE uploads/screenshots/ and use rsync --link-dest
  instead of cp -al. The per-device *_latest.jpg screenshots are rewritten
  24/7; cp -al aborts when a file mutates mid-copy and the prior script
  swallowed the error with 2>/dev/null, silently breaking content snapshots
  for ~8 weeks. rsync --link-dest hard-links unchanged files but tolerates
  in-flight changes; errors now go to backup.log.
- Retention sorts by NAME, not mtime: rsync -a / cp -al preserve the source
  dir's (frozen) mtime, so ls -dt treated fresh snapshots as oldest and pruned
  them. The timestamp is in the dir name, so name-sort is chronological.

README Backups section documents the cron setup + env knobs. Verified on prod.
2026-06-09 19:53:09 -05:00

80 lines
3.7 KiB
Bash
Executable file

#!/bin/bash
# ScreenTinker backup - nightly DB + content backup with point-in-time history.
#
# Install (self-hosters):
# 1. Set SCREENTINKER_DIR if your install isn't at /opt/screentinker.
# 2. Add a root (or service-user) cron entry, e.g.:
# 0 3 * * * /opt/screentinker/scripts/backup.sh
# 3. Restore with: sqlite3 .backup files copy straight back;
# cp -a backups/content-<ts>/<file> server/uploads/<file>
#
# What it keeps in $BACKUP_DIR:
# remote_display-<ts>.db atomic SQLite snapshot (kept $DB_KEEP_DAYS days)
# content-latest/ live mirror of uploads/ (current-state recovery)
# content-<ts>/ daily point-in-time copy (kept newest $DAILY_KEEP)
# content-monthly-<YYYYMM>/ long-horizon keep (kept newest $MONTHLY_KEEP)
# backup.log run log (errors are recorded, not swallowed)
#
# Design notes (learned the hard way):
# - Content snapshots EXCLUDE uploads/screenshots/: per-device *_latest.jpg files
# are rewritten 24/7, and `cp -al` aborts when a file mutates mid-copy. That race
# silently broke snapshots in one deployment for ~8 weeks. rsync --link-dest below
# hard-links unchanged files (cheap, like cp -al) but tolerates in-flight changes.
# - Retention sorts by NAME, not mtime: rsync -a / cp -al preserve the source dir's
# (often frozen) mtime, so `ls -dt` would treat a fresh snapshot as oldest and
# prune it. The timestamp is in the dir name, so name-sort is chronological.
set -o pipefail
APP_DIR="${SCREENTINKER_DIR:-/opt/screentinker}"
BACKUP_DIR="${BACKUP_DIR:-$APP_DIR/backups}"
DB="${DB:-$APP_DIR/server/db/remote_display.db}"
UPLOADS="${UPLOADS:-$APP_DIR/server/uploads}"
DAILY_KEEP="${DAILY_KEEP:-7}"
MONTHLY_KEEP="${MONTHLY_KEEP:-12}"
DB_KEEP_DAYS="${DB_KEEP_DAYS:-30}"
LOG="$BACKUP_DIR/backup.log"
TIMESTAMP="$(date +%Y%m%d-%H%M)"
MONTH="$(date +%Y%m)"
mkdir -p "$BACKUP_DIR"
log() { echo "$(date '+%F %T') $*" >> "$LOG"; }
log "=== backup start $TIMESTAMP (app=$APP_DIR) ==="
# 1) Atomic DB backup (safe while the server is running).
if sqlite3 "$DB" ".backup '$BACKUP_DIR/remote_display-$TIMESTAMP.db'" 2>>"$LOG"; then
log "db backup ok: remote_display-$TIMESTAMP.db"
else
log "ERROR: db backup failed (exit $?)"
fi
# 2) Refresh the live content mirror (full, incl. screenshots - current-state DR).
if rsync -a --delete "$UPLOADS/" "$BACKUP_DIR/content-latest/" 2>>"$LOG"; then
touch "$BACKUP_DIR/content-latest" # rsync -a leaves the dir mtime frozen; correct it
log "content-latest mirror refreshed"
else
log "ERROR: content-latest rsync failed (exit $?)"
fi
# 3) Point-in-time content snapshot (hard-linked, screenshots excluded - see notes).
SNAP="$BACKUP_DIR/content-$TIMESTAMP"
rsync -a --link-dest="$BACKUP_DIR/content-latest" --exclude='/screenshots/' \
"$BACKUP_DIR/content-latest/" "$SNAP/" 2>>"$LOG"
rc=$?
if { [ "$rc" -eq 0 ] || [ "$rc" -eq 24 ]; } && [ -d "$SNAP" ]; then
touch "$SNAP"
log "content snapshot ok: content-$TIMESTAMP ($(find "$SNAP" -type f | wc -l) files, rc=$rc)"
MONTHLY="$BACKUP_DIR/content-monthly-$MONTH"
if [ ! -d "$MONTHLY" ]; then
cp -al "$SNAP" "$MONTHLY" 2>>"$LOG" && touch "$MONTHLY" && log "monthly keep created: content-monthly-$MONTH"
fi
else
log "ERROR: content snapshot failed (rsync exit $rc) - see above"
fi
# 4) Retention (name-sorted = chronological; see notes).
find "$BACKUP_DIR" -maxdepth 1 -name "remote_display-*.db" -mtime +"$DB_KEEP_DAYS" -delete 2>>"$LOG"
ls -d "$BACKUP_DIR"/content-2* 2>/dev/null | sort | head -n -"$DAILY_KEEP" | xargs -r rm -rf 2>>"$LOG"
ls -d "$BACKUP_DIR"/content-monthly-* 2>/dev/null | sort | head -n -"$MONTHLY_KEEP" | xargs -r rm -rf 2>>"$LOG"
log "=== backup done $TIMESTAMP ==="