Files
vitya b332af66e4 fix(vds-backup): MSSQL backup INIT->FORMAT + estate backup audit
VDS daily backup пал с 12.06 (line 108, exit 1): mssql-блок (added 11.06,
commit f8ca0794) использовал `WITH ... INIT`, упирался в компрессованный
media-header майских .bak (созданы WITH COMPRESSION на Developer-источнике).
Express не пишет в compression-форматированный media set -> Msg 1844, молчаливый
провал первым же cron-запуском. 5 боевых CMS-баз 3 недели без offsite-копии.

Fix: INIT -> FORMAT (всегда новый media set, иммунно к остаткам). Прогон
verified зелёным: VDS backup OK 95m41s, все 5 .bak на kreknin
(MoreThenCms 910M, StayerCalculator 528M, StayerPrice 39M, TireService 4.5M,
stostayer 990M), speedup 22.62 (--link-dest хардлинкует).

- scripts/vds-backup-rsync-kreknin/run.sh: синхронизирован с задеплоенным
  (mssql-блока в репо не было); FORMAT
- .wiki/concepts/mssql-on-vds.md: gotcha #2 (INIT vs FORMAT) + backup-gap
- .wiki/concepts/backup-inventory-2026-06.md: новая — карта estate × что реально
  бэкапится (с доказательством); триаж дыр
- openwrt UCI backup настроен (cron 03:30 -> kreknin, restricted forced-command
  key) — документация в entity + inventory
- .tasks/kreknin-self-backup.md: backlog #1 SPOF (приёмник сам не бэкапится)
- STATUS.md: incident + audit summary

Урок: бэкап-шаг не готов, пока не предъявлен лог одного реального успеха;
прод-крон не должен быть первым тестом бэкап-пути.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-12 10:21:44 +03:00

188 lines
6.0 KiB
Bash

#!/bin/bash
# Daily VDS → kreknin backup. Triggered by /etc/cron.d/vds-backup at 05:00 MSK.
set -euo pipefail
ENV_FILE=/opt/stacks/backup/.env
NTFY_ENV=/opt/stacks/ntfy/.env
# shellcheck disable=SC1090
source "$ENV_FILE"
# shellcheck disable=SC1090
source "$NTFY_ENV"
TODAY=$(date +%Y-%m-%d)
LOG_DIR=/var/log/vds-backup
LOG="$LOG_DIR/$TODAY.log"
LOCK=/tmp/vds-backup.lock
DUMP_DIR=/tmp/vds-backup-dumps/$TODAY
DEST_PATH="$DEST_BASE/$TODAY"
LATEST="$DEST_BASE/latest"
mkdir -p "$LOG_DIR" "$DUMP_DIR"
exec >> "$LOG" 2>&1
echo "===== VDS backup start $TODAY $(date -Is) ====="
# single-instance lock
exec 9>"$LOCK"
if ! flock -n 9; then echo "already running, exiting"; exit 0; fi
START=$(date +%s)
SSH_OPTS="-i $SSH_KEY -o StrictHostKeyChecking=accept-new -o ConnectTimeout=30"
HOST_LABEL=VDS
SOURCE_DISPLAY="vds.kzntsv.site (89.253.255.94)"
notify() {
local title="$1" msg="$2" tags="${3:-}" prio="${4:-default}"
curl -sS --max-time 10 -u "$NTFY_USER:$NTFY_PASS" \
-H "Title: $title" -H "Tags: $tags" -H "Priority: $prio" \
-d "$msg" "$NTFY_URL/vds-backup" > /dev/null || true
}
email_send() {
local subj="$1" body="$2"
printf "From: noreply@snolla.com\nTo: %s\nSubject: %s\n\n%s\n" \
"$EMAIL" "$subj" "$body" | msmtp "$EMAIL" || true
}
# Duration human-format: Xm YYs (e.g. "21m07s")
fmt_duration() {
local s=$1
printf "%dm%02ds" $((s/60)) $((s%60))
}
on_error() {
local line=$1 rc=$2
local err="line $line, exit $rc"
local now=$(date +%s)
local dur=$((now - ${START:-now}))
local dur_h=$(fmt_duration "$dur")
local tail_log
tail_log=$(tail -40 "$LOG" 2>/dev/null || true)
echo "FATAL: $HOST_LABEL backup FAILED — $err"
# ntfy push (single line)
notify "$HOST_LABEL backup FAILED $TODAY" \
"After $dur_h: $err. See $LOG" \
red_circle high
# email (structured)
email_send "[$HOST_LABEL] backup FAILED $TODAY" \
"$HOST_LABEL daily backup FAILED.
Date: $TODAY
Duration: $dur_h
Error: $err
Source: $SOURCE_DISPLAY
Log: $LOG
Tail (last 40 lines):
$tail_log"
exit "$rc"
}
trap 'on_error $LINENO $?' ERR
# === Step 1: DB dumps ===
echo "--- DB dumps to $DUMP_DIR ---"
docker exec postgres pg_dumpall -U postgres | gzip > "$DUMP_DIR/postgres.sql.gz"
echo "postgres: $(stat -c %s "$DUMP_DIR/postgres.sql.gz") bytes"
docker exec mariadb mariadb-dump -uroot -p"$MARIADB_ROOT_PASSWORD" --all-databases --single-transaction --quick 2>/dev/null | gzip > "$DUMP_DIR/mariadb.sql.gz"
echo "mariadb: $(stat -c %s "$DUMP_DIR/mariadb.sql.gz") bytes"
docker exec mongo mongodump --archive --quiet --ssl --sslAllowInvalidCertificates --uri="mongodb://${MONGO_INITDB_ROOT_USERNAME}:${MONGO_INITDB_ROOT_PASSWORD}@localhost:27017/?authSource=admin" 2>/dev/null | gzip > "$DUMP_DIR/mongo.archive.gz"
echo "mongo: $(stat -c %s "$DUMP_DIR/mongo.archive.gz") bytes"
docker exec redis redis-cli --tls --insecure -a "$REDIS_PASSWORD" --no-auth-warning SAVE > /dev/null
docker cp redis:/data/dump.rdb "$DUMP_DIR/redis.rdb"
gzip -f "$DUMP_DIR/redis.rdb"
echo "redis: $(stat -c %s "$DUMP_DIR/redis.rdb.gz") bytes"
# MSSQL: backup 5 prod DBs via sqlcmd into bind-mounted /var/opt/mssql/backups/
# WITH FORMAT (not INIT): Express can't write to a media set previously formatted
# WITH COMPRESSION (e.g. migration-era .bak); FORMAT recreates the media set fresh.
source /opt/stacks/databases/mssql/.env # injects MSSQL_SA_PASSWORD
MSSQL_BACKUP_DIR=/opt/stacks/databases/mssql/backups
for DB in MoreThenCms StayerCalculator StayerPrice TireService stostayer; do
docker exec mssql /opt/mssql-tools18/bin/sqlcmd \
-S localhost -U sa -P "$MSSQL_SA_PASSWORD" -C -b \
-Q "BACKUP DATABASE [$DB] TO DISK='/var/opt/mssql/backups/${DB}.bak' WITH COPY_ONLY, FORMAT" \
> /dev/null
mv "$MSSQL_BACKUP_DIR/${DB}.bak" "$DUMP_DIR/mssql-${DB}.bak"
echo "mssql $DB: $(stat -c %s "$DUMP_DIR/mssql-${DB}.bak") bytes"
done
unset MSSQL_SA_PASSWORD
# === Step 2: rsync to kreknin ===
echo "--- rsync to $DEST_USER@$DEST_HOST:$DEST_PATH ---"
ssh $SSH_OPTS "$DEST_USER@$DEST_HOST" "mkdir -p '$DEST_PATH'"
LINK_DEST_ARG=""
if ssh $SSH_OPTS "$DEST_USER@$DEST_HOST" "test -L '$LATEST' || test -d '$LATEST'" 2>/dev/null; then
LINK_DEST_ARG="--link-dest=$LATEST"
fi
rsync -aHh --info=stats2 --delete $LINK_DEST_ARG \
-e "ssh $SSH_OPTS" \
/opt/stacks/gitea \
/opt/stacks/verdaccio/storage \
/opt/stacks/verdaccio/config \
/opt/stacks/registry \
/opt/stacks/traefik \
/opt/stacks/portainer \
/opt/stacks/ntfy \
/opt/stacks/backup \
/opt/stacks/owncloud \
/etc/ssh \
/etc/ufw \
/etc/hosts \
/etc/docker \
"$DUMP_DIR" \
/home/vitya/.ssh \
"$DEST_USER@$DEST_HOST:$DEST_PATH/" || true
# Update latest symlink
ssh $SSH_OPTS "$DEST_USER@$DEST_HOST" "rm -f '$LATEST' && ln -s '$DEST_PATH' '$LATEST'"
# === Step 3: retention prune ===
echo "--- prune snapshots older than $RETENTION_DAYS days ---"
ssh $SSH_OPTS "$DEST_USER@$DEST_HOST" "cd '$DEST_BASE' && ls -1d 20*-*-* 2>/dev/null | sort | head -n -$RETENTION_DAYS | xargs -r rm -rf"
# === Step 4: local cleanup ===
rm -rf "$DUMP_DIR"
# === Step 5: report ===
END=$(date +%s)
DURATION=$((END - START))
SIZE=$(ssh $SSH_OPTS "$DEST_USER@$DEST_HOST" "du -sh '$DEST_PATH' 2>/dev/null | cut -f1" || echo "?")
SNAPSHOT_COUNT=$(ssh $SSH_OPTS "$DEST_USER@$DEST_HOST" "ls -1d $DEST_BASE/20*-*-* 2>/dev/null | wc -l" || echo "?")
DURATION_HUMAN=$(fmt_duration "$DURATION")
# ntfy push (single line)
NTFY_BODY="$DURATION_HUMAN, size=$SIZE, snapshots=$SNAPSHOT_COUNT, dest=kreknin:$DEST_PATH"
notify "$HOST_LABEL backup OK $TODAY" "$NTFY_BODY" green_circle
# email (structured)
EMAIL_BODY="$HOST_LABEL daily backup completed successfully.
Date: $TODAY
Duration: $DURATION_HUMAN
Size: $SIZE
Snapshots: $SNAPSHOT_COUNT
Source: $SOURCE_DISPLAY
Dest: kreknin:$DEST_PATH
Components:
- /opt/stacks/{gitea,verdaccio,registry,traefik,portainer,ntfy,backup,owncloud}
- /etc/{ssh,ufw,hosts,docker}
- /home/vitya/.ssh
- DB dumps: postgres, mariadb, mongo, redis
Log: $LOG"
email_send "[$HOST_LABEL] backup OK $TODAY" "$EMAIL_BODY"
echo "$HOST_LABEL backup OK: $NTFY_BODY"
echo "===== $HOST_LABEL backup done $(date -Is) ====="