🔄 第33课：灾难恢复

velero install --provider aws --bucket my-velero-backup \
  --backup-location-config region=ap-southeast-1 \
  --snapshot-location-config region=ap-southeast-1 --use-node-agent

# 定时备份(每天2点)
velero schedule create daily-backup --schedule "0 2 * * *" \
  --include-namespaces production --ttl 720h

# 手动备份
velero backup create manual-backup --include-namespaces production

# 查看备份状态
velero backup get

# 恢复备份
velero restore create --from-backup daily-backup
velero restore get

#!/bin/bash
# pg-backup.sh
PG_HOST="${PG_HOST:-localhost}"
S3_BUCKET="s3://my-db-backups"
DATE=$(date +%Y%m%d_%H%M%S)

# 全量逻辑备份
pg_dumpall -h $PG_HOST -U postgres | gzip > /tmp/pg_all_$DATE.sql.gz
aws s3 cp /tmp/pg_all_$DATE.sql.gz $S3_BUCKET/postgresql/
rm -f /tmp/pg_all_$DATE.sql.gz

# 清理30天前备份
aws s3 ls $S3_BUCKET/postgresql/ | while read -r _ _ _ f; do
    file_date=$(echo "$f" | grep -oP '\d{8}')
    diff=$(( ($(date +%s) - $(date -d "$file_date" +%s)) / 86400 ))
    [[ $diff -gt 30 ]] && aws s3 rm $S3_BUCKET/postgresql/$f
done

# 恢复
gunzip -c pg_all_20260519_020000.sql.gz | psql -h $PG_HOST -U postgres

#!/bin/bash
# dr-drill.sh
START_TIME=$(date +%s)
LOG="/var/log/dr-drill-$(date +%Y%m%d).log"

log() { echo "[$(date '+%H:%M:%S')] $*" | tee -a "$LOG"; }

log "🚨 DR演练开始"

# 验证备份
velero backup get | grep "Completed" || { log "❌ 无可用备份"; exit 1; }

# 恢复到DR环境
velero restore create dr-drill-restore --from-backup daily-backup

for i in $(seq 1 60); do
    STATUS=$(velero restore get dr-drill-restore -o jsonpath='{.status.phase}')
    [[ "$STATUS" == "Completed" ]] && break
    log "等待恢复... ($i/60) status=$STATUS"
    sleep 10
done

RESTORE_TIME=$(( $(date +%s) - START_TIME ))
log "恢复耗时: ${RESTORE_TIME}秒 (RTO目标: < 4小时)"

# 验证服务
for svc in api-server web-app order-service; do
    if curl -sf "http://$svc.dr-env.svc/health" > /dev/null 2>&1; then
        log "✅ $svc 健康"
    else
        log "❌ $svc 不可用"
    fi
done

log "DR演练完成"