stemedb/scripts/backup-stemedb.sh

#!/usr/bin/env bash
#
# StemeDB Backup Script
#
# Creates a timestamped backup of WAL and database files.
#
# Usage:
#   ./scripts/backup-stemedb.sh                     # Default backup to backups/
#   ./scripts/backup-stemedb.sh --output /mnt/nfs   # Custom output directory
#   ./scripts/backup-stemedb.sh --wal-only           # Backup WAL only (faster)
#
# Exit codes:
#   0 - Backup completed successfully
#   1 - Backup failed
#

set -euo pipefail

# Configuration
readonly SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
readonly PROJECT_DIR="$(dirname "$SCRIPT_DIR")"
readonly WAL_DIR="${STEMEDB_WAL_DIR:-${PROJECT_DIR}/data/wal}"
readonly DB_DIR="${STEMEDB_DB_DIR:-${PROJECT_DIR}/data/db}"
readonly TIMESTAMP="$(date +%Y%m%d-%H%M%S)"

# Colors (if terminal supports it)
if [[ -t 1 ]]; then
    RED='\033[0;31m'
    GREEN='\033[0;32m'
    YELLOW='\033[0;33m'
    BLUE='\033[0;34m'
    NC='\033[0m'
else
    RED=''
    GREEN=''
    YELLOW=''
    BLUE=''
    NC=''
fi

# Logging helpers
info() { echo -e "${BLUE}[INFO]${NC} $*"; }
success() { echo -e "${GREEN}[OK]${NC} $*"; }
warn() { echo -e "${YELLOW}[WARN]${NC} $*"; }
fail() { echo -e "${RED}[FAIL]${NC} $*"; exit 1; }

# Defaults
OUTPUT_DIR="${PROJECT_DIR}/backups"
WAL_ONLY=false

# Parse arguments
while [[ $# -gt 0 ]]; do
    case $1 in
        --output)
            OUTPUT_DIR="$2"
            shift 2
            ;;
        --wal-only)
            WAL_ONLY=true
            shift
            ;;
        --help|-h)
            echo "Usage: $0 [--output <dir>] [--wal-only]"
            echo ""
            echo "Create a timestamped backup of StemeDB data."
            echo ""
            echo "Options:"
            echo "  --output <dir>   Output directory (default: backups/)"
            echo "  --wal-only       Backup WAL directory only (skip DB)"
            echo "  --help           Show this help message"
            echo ""
            echo "Environment:"
            echo "  STEMEDB_WAL_DIR  WAL directory (default: data/wal)"
            echo "  STEMEDB_DB_DIR   Database directory (default: data/db)"
            exit 0
            ;;
        *)
            fail "Unknown argument: $1 (use --help for usage)"
            ;;
    esac
done

readonly BACKUP_DIR="${OUTPUT_DIR}/stemedb-backup-${TIMESTAMP}"

# Cleanup partial backup on failure
cleanup() {
    local exit_code=$?
    if [[ $exit_code -ne 0 && -d "$BACKUP_DIR" ]]; then
        warn "Backup failed, removing partial backup at ${BACKUP_DIR}"
        rm -rf "$BACKUP_DIR"
    fi
}
trap cleanup EXIT

main() {
    echo ""
    echo "=========================================="
    echo "  StemeDB Backup"
    echo "=========================================="
    echo ""

    # Validate source directories
    if [[ ! -d "$WAL_DIR" ]]; then
        fail "WAL directory not found: ${WAL_DIR}"
    fi

    if [[ -z "$(ls -A "$WAL_DIR" 2>/dev/null)" ]]; then
        fail "WAL directory is empty: ${WAL_DIR}"
    fi

    if [[ "$WAL_ONLY" == "false" ]]; then
        if [[ ! -d "$DB_DIR" ]]; then
            fail "DB directory not found: ${DB_DIR}"
        fi
        if [[ -z "$(ls -A "$DB_DIR" 2>/dev/null)" ]]; then
            fail "DB directory is empty: ${DB_DIR}"
        fi
    fi

    # Create backup directory
    mkdir -p "$BACKUP_DIR"
    info "Backup directory: ${BACKUP_DIR}"

    # Backup WAL (append-only, safe to copy live)
    info "Copying WAL directory..."
    rsync -a "${WAL_DIR}/" "${BACKUP_DIR}/wal/"
    local wal_files
    wal_files=$(find "${BACKUP_DIR}/wal" -type f | wc -l)
    local wal_size
    wal_size=$(du -sh "${BACKUP_DIR}/wal" | cut -f1)
    success "WAL: ${wal_files} files, ${wal_size}"

    # Backup DB (unless --wal-only)
    local db_files=0
    local db_size="0"
    if [[ "$WAL_ONLY" == "false" ]]; then
        info "Copying DB directory..."
        rsync -a "${DB_DIR}/" "${BACKUP_DIR}/db/"
        db_files=$(find "${BACKUP_DIR}/db" -type f | wc -l)
        db_size=$(du -sh "${BACKUP_DIR}/db" | cut -f1)
        success "DB: ${db_files} files, ${db_size}"
    else
        info "Skipping DB (--wal-only)"
    fi

    # Compute total size
    local total_size
    total_size=$(du -sh "$BACKUP_DIR" | cut -f1)

    # Write metadata
    cat > "${BACKUP_DIR}/backup-metadata.json" <<METADATA
{
  "timestamp": "$(date -u +%Y-%m-%dT%H:%M:%SZ)",
  "source_wal_dir": "${WAL_DIR}",
  "source_db_dir": "${DB_DIR}",
  "wal_only": ${WAL_ONLY},
  "wal_file_count": ${wal_files},
  "db_file_count": ${db_files},
  "total_size": "${total_size}",
  "hostname": "$(hostname)",
  "stemedb_version": "$(cargo metadata --format-version=1 --no-deps 2>/dev/null | grep -o '"stemedb-api","version":"[^"]*"' | head -1 | cut -d'"' -f6 || echo "unknown")"
}
METADATA
    success "Metadata written"

    # Summary
    echo ""
    echo "=========================================="
    echo -e "  ${GREEN}Backup complete${NC}"
    echo "=========================================="
    echo ""
    echo "  Location:  ${BACKUP_DIR}"
    echo "  WAL files: ${wal_files} (${wal_size})"
    if [[ "$WAL_ONLY" == "false" ]]; then
        echo "  DB files:  ${db_files} (${db_size})"
    fi
    echo "  Total:     ${total_size}"
    echo ""
    echo "Restore with:"
    echo "  ./scripts/restore-stemedb.sh ${BACKUP_DIR}"
    echo ""
}

main "$@"