文件
hyb-oa/scripts/qa/run-backup-restore-drill.sh
T

151 行
6.2 KiB
Bash
可执行文件

#!/usr/bin/env bash
set -euo pipefail
ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)"
DRILL_RUN_ID="drill-$(date -u +%Y%m%dT%H%M%SZ)-$$"
CONTAINER_NAME="hyb-oa-backup-drill-$$"
DRILL_DIR="$(mktemp -d)"
POSTGRES_USER="drill_user"
POSTGRES_PASSWORD="drill_password"
POSTGRES_DB="hyb_oa_restore_shadow"
cleanup() {
docker rm -f "$CONTAINER_NAME" >/dev/null 2>&1 || true
if [[ -n "${DRILL_DIR:-}" && -d "$DRILL_DIR" && "$DRILL_DIR" == /tmp/* ]]; then
rm -rf -- "$DRILL_DIR"
fi
}
trap cleanup EXIT
require_command() {
if ! command -v "$1" >/dev/null 2>&1; then
echo "[ERR] required command missing: $1" >&2
exit 1
fi
}
for command_name in docker pg_dump pg_restore psql tar sha256sum node; do
require_command "$command_name"
done
START_EPOCH="$(date +%s)"
SOURCE_STORAGE="$DRILL_DIR/source-storage"
RESTORED_STORAGE="$DRILL_DIR/restored-storage"
NONEMPTY_STORAGE="$DRILL_DIR/nonempty-storage"
BACKUP_ROOT="$DRILL_DIR/backups"
mkdir -p "$SOURCE_STORAGE/nested" "$NONEMPTY_STORAGE" "$BACKUP_ROOT"
printf 'synthetic attachment for restore drill\n' >"$SOURCE_STORAGE/nested/marker.txt"
printf 'must not be overwritten\n' >"$NONEMPTY_STORAGE/existing.txt"
docker run -d --rm --name "$CONTAINER_NAME" \
-e "POSTGRES_USER=$POSTGRES_USER" \
-e "POSTGRES_PASSWORD=$POSTGRES_PASSWORD" \
-e "POSTGRES_DB=$POSTGRES_DB" \
-p 127.0.0.1::5432 postgres:16-alpine >/dev/null
for _ in $(seq 1 60); do
if docker exec "$CONTAINER_NAME" pg_isready -U "$POSTGRES_USER" -d "$POSTGRES_DB" >/dev/null 2>&1; then
break
fi
sleep 1
done
docker exec "$CONTAINER_NAME" pg_isready -U "$POSTGRES_USER" -d "$POSTGRES_DB" >/dev/null
HOST_PORT="$(docker port "$CONTAINER_NAME" 5432/tcp | awk -F: 'NR==1 {print $NF}')"
DATABASE_URL="postgresql://$POSTGRES_USER:$POSTGRES_PASSWORD@127.0.0.1:$HOST_PORT/$POSTGRES_DB"
export DATABASE_URL STORAGE_DIR_HOST="$SOURCE_STORAGE" BACKUP_DIR="$BACKUP_ROOT" BACKUP_RUN_ID="$DRILL_RUN_ID"
export DB_TOOL_MODE="docker" DB_CONTAINER="$CONTAINER_NAME" POSTGRES_USER POSTGRES_PASSWORD POSTGRES_DB
psql "$DATABASE_URL" -v ON_ERROR_STOP=1 -qc 'CREATE TABLE restore_drill_marker (value text NOT NULL)'
psql "$DATABASE_URL" -v ON_ERROR_STOP=1 -qc "INSERT INTO restore_drill_marker(value) VALUES ('synthetic-round-trip')"
"$ROOT_DIR/scripts/backup-all.sh"
BATCH_DIR="$BACKUP_ROOT/hyb-oa-backup-$DRILL_RUN_ID"
DB_BACKUP="$BATCH_DIR/database.dump"
STORAGE_BACKUP="$BATCH_DIR/storage.tar.gz"
node - "$BATCH_DIR/manifest.json" "$DB_BACKUP" "$STORAGE_BACKUP" <<'NODE'
const crypto = require("node:crypto");
const fs = require("node:fs");
const path = require("node:path");
const [manifestPath, dbPath, storagePath] = process.argv.slice(2);
const manifestText = fs.readFileSync(manifestPath, "utf8");
const manifest = JSON.parse(manifestText);
if (manifest.status !== "complete" || manifest.schemaVersion !== 1) throw new Error("manifest contract mismatch");
if (path.isAbsolute(manifest.artifacts.database.file) || path.isAbsolute(manifest.artifacts.storage.file)) {
throw new Error("manifest contains absolute artifact path");
}
if (/postgresql:|DATABASE_URL|drill_password/.test(manifestText)) throw new Error("manifest contains sensitive connection data");
const sha256 = (file) => crypto.createHash("sha256").update(fs.readFileSync(file)).digest("hex");
if (sha256(dbPath) !== manifest.artifacts.database.sha256) throw new Error("database checksum mismatch");
if (sha256(storagePath) !== manifest.artifacts.storage.sha256) throw new Error("storage checksum mismatch");
NODE
psql "$DATABASE_URL" -v ON_ERROR_STOP=1 -qc 'DROP TABLE restore_drill_marker'
docker exec -e PGPASSWORD="$POSTGRES_PASSWORD" "$CONTAINER_NAME" \
createdb -U "$POSTGRES_USER" hyb_oa_prod
if DATABASE_URL="postgresql://$POSTGRES_USER:$POSTGRES_PASSWORD@127.0.0.1:$HOST_PORT/hyb_oa_prod" \
POSTGRES_DB=hyb_oa_prod RESTORE_SCOPE=shadow RESTORE_CONFIRM_DATABASE=hyb_oa_prod \
"$ROOT_DIR/scripts/restore-db.sh" "$DB_BACKUP" >/dev/null 2>&1; then
echo "[ERR] production-like DB name was accepted as a shadow target" >&2
exit 1
fi
CORRUPT_DB="$DRILL_DIR/corrupt.dump"
head -c 32 "$DB_BACKUP" >"$CORRUPT_DB"
if RESTORE_SCOPE=shadow RESTORE_CONFIRM_DATABASE="$POSTGRES_DB" \
"$ROOT_DIR/scripts/restore-db.sh" "$CORRUPT_DB" >/dev/null 2>&1; then
echo "[ERR] corrupt DB dump was accepted" >&2
exit 1
fi
if RESTORE_SCOPE=shadow RESTORE_CONFIRM_DATABASE=wrong_target \
"$ROOT_DIR/scripts/restore-db.sh" "$DB_BACKUP" >/dev/null 2>&1; then
echo "[ERR] mismatched DB confirmation was accepted" >&2
exit 1
fi
RESTORE_SCOPE=shadow RESTORE_CONFIRM_DATABASE="$POSTGRES_DB" \
"$ROOT_DIR/scripts/restore-db.sh" "$DB_BACKUP"
RESTORED_VALUE="$(psql "$DATABASE_URL" -Atqc 'SELECT value FROM restore_drill_marker')"
if [[ "$RESTORED_VALUE" != "synthetic-round-trip" ]]; then
echo "[ERR] DB round-trip marker mismatch" >&2
exit 1
fi
if RESTORE_SCOPE=shadow "$ROOT_DIR/scripts/restore-storage.sh" "$STORAGE_BACKUP" "$NONEMPTY_STORAGE" >/dev/null 2>&1; then
echo "[ERR] non-empty storage target was accepted" >&2
exit 1
fi
if RESTORE_SCOPE=shadow "$ROOT_DIR/scripts/restore-storage.sh" \
"$STORAGE_BACKUP" "$ROOT_DIR/deploy/trialrun/data/storage" >/dev/null 2>&1; then
echo "[ERR] default live storage path was accepted as a shadow target" >&2
exit 1
fi
RESTORE_SCOPE=shadow "$ROOT_DIR/scripts/restore-storage.sh" "$STORAGE_BACKUP" "$RESTORED_STORAGE"
cmp "$SOURCE_STORAGE/nested/marker.txt" "$RESTORED_STORAGE/nested/marker.txt"
CORRUPT_TAR="$DRILL_DIR/corrupt.tar.gz"
head -c 32 "$STORAGE_BACKUP" >"$CORRUPT_TAR"
if RESTORE_SCOPE=shadow "$ROOT_DIR/scripts/restore-storage.sh" "$CORRUPT_TAR" "$DRILL_DIR/corrupt-target" >/dev/null 2>&1; then
echo "[ERR] corrupt storage archive was accepted" >&2
exit 1
fi
DANGEROUS_TAR="$DRILL_DIR/dangerous.tar.gz"
tar -czf "$DANGEROUS_TAR" --transform='s|nested/marker.txt|../escape.txt|' \
-C "$SOURCE_STORAGE" nested/marker.txt
if RESTORE_SCOPE=shadow "$ROOT_DIR/scripts/restore-storage.sh" "$DANGEROUS_TAR" "$DRILL_DIR/dangerous-target" >/dev/null 2>&1; then
echo "[ERR] dangerous storage archive path was accepted" >&2
exit 1
fi
END_EPOCH="$(date +%s)"
ELAPSED_SECONDS="$((END_EPOCH - START_EPOCH))"
echo "[PASS] isolated backup/restore drill completed"
echo "[INFO] local synthetic RTO seconds: $ELAPSED_SECONDS"