# 保留策略
1. 短期保留
– 日备份保留1周
– 用于快速恢复

2. 中期保留
– 周备份保留1个月
– 用于中期恢复

3. 长期保留
– 月备份保留1年
– 用于合规审计

# HDFS备份
# 1. DistCp备份
hadoop distcp /user /backup/user_$(date +%Y%m%d)

# 2. 快照备份
hdfs dfsadmin -allowSnapshot /user
hdfs dfs -createSnapshot /user user_snapshot_$(date +%Y%m%d)

# 3. 查看快照
hdfs lsSnapshottableDir
hdfs dfs -ls /user/.snapshot/

# 4. 备份脚本
#!/bin/bash
# backup_hdfs.sh
# from:www.itpux.com.qq113257174.wx:itpux-com
# web: http://www.fgedu.net.cn

DATE=$(date +%Y%m%d_%H%M%S)
BACKUP_DIR=/backup/hdfs
LOG_DIR=/backup/logs
mkdir -p $BACKUP_DIR $LOG_DIR
LOG_FILE=$LOG_DIR/backup_$DATE.log

echo “Start backup at $(date)” >> $LOG_FILE

hadoop distcp -update -delete /user $BACKUP_DIR/user_$DATE >> $LOG_FILE 2>&1

if [ $? -eq 0 ]; then
echo “Backup successful at $(date)” >> $LOG_FILE
else
echo “Backup failed at $(date)” >> $LOG_FILE
fi

# 添加到crontab
# 0 2 * * * /bigdata/scripts/backup_hdfs.sh

# Hive备份
# 1. 元数据备份
# 备份MySQL
mysqldump -u root -pfgedu123 metastore > /backup/hive/metastore_$(date +%Y%m%d).sql

# 2. 数据备份
hadoop distcp /user/hive/warehouse /backup/hive/warehouse_$(date +%Y%m%d)

# 3. 导出表
hive -e “EXPORT TABLE fgedudb.fgedu_table TO ‘/backup/hive/fgedu_table_$(date +%Y%m%d)'”

# 4. 备份脚本
#!/bin/bash
# backup_hive.sh
DATE=$(date +%Y%m%d)
BACKUP_DIR=/backup/hive
mkdir -p $BACKUP_DIR

# 备份元数据
mysqldump -u root -pfgedu123 metastore > $BACKUP_DIR/metastore_$DATE.sql

# 备份数据
hadoop distcp /user/hive/warehouse $BACKUP_DIR/warehouse_$DATE

# HBase备份
# 1. 快照备份
hbase snapshot create_snapshot ‘fgedu_table’, ‘fgedu_table_snapshot_$(date +%Y%m%d)’

# 2. 查看快照
hbase list_snapshots

# 3. 导出快照
hbase snapshot export_snapshot -snapshot fgedu_table_snapshot_20240408 -copy-to hdfs://backup-cluster/backup/hbase/

# 4. Export工具
hbase org.apache.hadoop.hbase.mapreduce.Export fgedudb.fgedu_table /backup/hbase/fgedu_table_$(date +%Y%m%d)

# 5. 备份脚本
#!/bin/bash
# backup_hbase.sh
DATE=$(date +%Y%m%d)
hbase snapshot create_snapshot ‘fgedu_table’, ‘fgedu_table_snapshot_$DATE’