bitscn.com
mysql cluster-备份恢复初步测试
sql节点上面录入数据：
 建立测试的数据库和表
create database bg;
create table bgt1 (id int,`name` varchar(20),primary key(`id`))engine=ndbcluster;
insert into bgt1 values(1,'zhang1'),(2,'zhang2'),(3,'zhang3'),(4,'zhang4');
8.2 管理节点上面，开始备份 
ndb_mgm> start backup
waiting for completed, this may take several minutes
node 4: backup 1 started from node 1
node 4: backup 1 started from node 1 completed
startgcp: 184725 stopgcp: 184772
#records: 722078 #logrecords: 0
data: 287345616 bytes log: 0 bytes
ndb_mgm>
8.3 再次插入几条数据(为了保持和正式环境尽可能接近，在插入数据中间穿插了flush logs操作!)
insert into bgt1 values(5,'zhang5'),(6,'zhang6'); 
flush logs;
insert into bgt1 values(7,'zhang7'),(8,'zhang8');
8.4 删掉sql节点的数据。
mysql> drop database bg;
query ok, 2 rows affected (6.16 sec)
8.5 关闭mysqld服务器。
[root@banggo data]# /etc/rc.d/init.d/mysqld stop
shutting down mysql......120718 18:58:11 mysqld_safe mysqld from pid file /usr/local/mysql/data/banggo.local.pid ended
  [确定]
[1]+  done                    /usr/local/mysql/bin/mysqld_safe
[root@banggo data]# 
[root@banggo data]# /etc/rc.d/init.d/mysqld stop
shutting down mysql....                                    [确定]
[root@banggo data]#
8.6 重新启动节点
ndb_mgm> shutdown
node 4: cluster shutdown initiated
node 4: node shutdown completed.
2 ndb cluster node(s) have shutdown.
disconnecting to allow management server to shutdown.
ndb_mgm> exit
[root@banggo mysql-cluster]# ndb_mgmd -f /usr/local/mysql/cluster-conf/config.ini --reload
mysql cluster management server mysql-5.5.19 ndb-7.2.4
8.7 重新启动数据节点
ndbd --initial
8.8 在数据节点上面进行恢复。
/home/mysql-cluster-gpl-7.2.4-linux2.6-x86_64/bin/ndb_restore -e  -c 10.100.200.36 -n 4 -b 1 -m --backup_path=/var/lib/mysql-cluster/backup/backup-1/
其中backup_path 在默认的数据节点的根目录下面(找了很久，一开始以为在配置文件里面)
第一步骤 -m操作
   [root@test-db-20053 backup-1]# /home/mysql-cluster-gpl-7.2.4-linux2.6-x86_64/bin/ndb_restore -e  -c 10.100.200.36 -n 4 -b 1 -m --backup_path=/var/lib/mysql-cluster/backup/backup-1/
nodeid = 4
backup id = 1
backup path = /var/lib/mysql-cluster/backup/backup-1/
opening file '/var/lib/mysql-cluster/backup/backup-1/backup-1.4.ctl'
file size 61160 bytes
backup version in files: ndb-6.3.11 ndb version: mysql-5.5.19 ndb-7.2.4
stop gcp of backup: 184771
connected to ndb!!
successfully restored table `test/def/t2`
successfully restored table event repl$test/t2
successfully restored table `bg/def/#sql-303d_2`
successfully restored table event repl$bg/#sql-303d_2
successfully restored table `bg/def/#sql-51f0_3`
successfully restored table event repl$bg/#sql-51f0_3
successfully restored table `test/def/t11`
successfully restored table event repl$test/t11
successfully restored table `ndb/def/ndborder_goods`
successfully restored table event repl$ndb/ndborder_goods
successfully restored table `bg/def/bgt1`
successfully restored table event repl$bg/bgt1
successfully restored table `test/def/ndborder_info_history`
successfully restored table event repl$test/ndborder_info_history
successfully restored table `mysql/def/ndb_schema`
successfully restored table event repl$mysql/ndb_schema
successfully restored table `mysql/def/ndb_apply_status`
successfully restored table event repl$mysql/ndb_apply_status
successfully restored table `ndb/def/ndbtest`
successfully restored table event repl$ndb/ndbtest
successfully created index `primary` on `ndborder_info_history`
successfully created index `uniq_order_os` on `ndborder_goods`
successfully created index `is_update` on `ndborder_info_history`
successfully created index `primary` on `#sql-51f0_3`
successfully created index `sku_sn` on `ndborder_goods`
successfully created index `primary` on `bgt1`
successfully created index `exchange_from` on `ndborder_goods`
successfully created index `addtime` on `ndborder_info_history`
successfully created index `relating_return_sn` on `ndborder_info_history`
successfully created index `primary` on `ndborder_goods`
successfully created index `order_from` on `ndborder_info_history`
successfully created index `order_out_sn` on `ndborder_info_history`
successfully created index `primary` on `#sql-303d_2`
successfully created index `order_status` on `ndborder_info_history`
successfully created index `user_id` on `ndborder_info_history`
successfully created index `uniq_order_os$unique` on `ndborder_goods`
successfully created index `order_sn` on `ndborder_goods`
ndbt_programexit: 0 - ok
[root@test-db-20053 backup-1]# 
8.9 第二步骤 -r操作(如果有n个node，则需要执行n次)
[root@test-db-20053 backup-1]# /home/mysql-cluster-gpl-7.2.4-linux2.6-x86_64/bin/ndb_restore -e  -c 10.100.200.36 -n 4 -b 1 -r --backup_path=/var/lib/mysql-cluster/backup/backup-1/
nodeid = 4
backup id = 1
backup path = /var/lib/mysql-cluster/backup/backup-1/
opening file '/var/lib/mysql-cluster/backup/backup-1/backup-1.4.ctl'
file size 61160 bytes
backup version in files: ndb-6.3.11 ndb version: mysql-5.5.19 ndb-7.2.4
stop gcp of backup: 184771
connected to ndb!!
opening file '/var/lib/mysql-cluster/backup/backup-1/backup-1-0.4.data'
file size 287834112 bytes
_____________________________________________________
processing data in table: test/def/t2(20) fragment 0
_____________________________________________________
processing data in table: bg/def/#sql-303d_2(34) fragment 0
_____________________________________________________
processing data in table: bg/def/#sql-51f0_3(32) fragment 0
_____________________________________________________
processing data in table: mysql/def/ndb$blob_7_3(8) fragment 0
_____________________________________________________
processing data in table: test/def/t11(11) fragment 0
_____________________________________________________
processing data in table: ndb/def/ndborder_goods(12) fragment 0
_____________________________________________________
processing data in table: ndb/def/ndb$blob_12_13(13) fragment 0
_____________________________________________________
processing data in table: mysql/def/ndb_index_stat_head(4) fragment 0
_____________________________________________________
processing data in table: sys/def/ndb$events_0(3) fragment 0
_____________________________________________________
processing data in table: sys/def/systab_0(2) fragment 0
_____________________________________________________
processing data in table: mysql/def/ndb_index_stat_sample(5) fragment 0
_____________________________________________________
processing data in table: bg/def/bgt1(36) fragment 0
_____________________________________________________
processing data in table: test/def/ndborder_info_history(21) fragment 0
temporary error: 1220: redo log files overloaded (increase fragmentlogfilesize)
temporary error: 1220: redo log files overloaded (increase fragmentlogfilesize)
temporary error: 1220: redo log files overloaded (increase fragmentlogfilesize)
temporary error: 1220: redo log files overloaded (increase fragmentlogfilesize)
temporary error: 1220: redo log files overloaded (increase fragmentlogfilesize)
temporary error: 1220: redo log files overloaded (increase fragmentlogfilesize)
temporary error: 1220: redo log files overloaded (increase fragmentlogfilesize)
temporary error: 1220: redo log files overloaded (increase fragmentlogfilesize)
temporary error: 1220: redo log files overloaded (increase fragmentlogfilesize)
temporary error: 1220: redo log files overloaded (increase fragmentlogfilesize)
在执行 -r操作的时候 报错
【ok】飞鸿大哥说是由于redo log 文件太小了要加大，不影响恢复效果。参考了http://bugs.mysql.com/bug.php?id=19651 这上面的人也这么讲。
mysql> show databases;
+--------------------+
| database           |
+--------------------+
| information_schema |
| mysql              |
| ndb                |
| ndbinfo            |
| performance_schema |
| test               |
+--------------------+
6 rows in set (0.00 sec)
-- 没有库？看来要重建了
mysql> create database bg; -- 重建
query ok, 1 row affected (0.06 sec)
mysql> use bg
database changed
mysql> show tables; -- ok，看下表
+--------------+
| tables_in_bg |
+--------------+
| bgt1         |
+--------------+
1 row in set (0.00 sec)
mysql> select * from bgt1; --数据恢复过来了
+----+--------+
| id | name   |
+----+--------+
|  3 | zhang3 |
|  1 | zhang1 |
|  2 | zhang2 |
|  4 | zhang4 |
+----+--------+
4 rows in set (0.01 sec)
8.10 找一个mysqld节点，在管理节点进入单用户模式，然后启动sql节点，启动该mysqld节点，并登陆找到最大的epoch的值
ndb_mgm> enter single user mode 10;
single user mode entered
access is granted for api node 10 only.
ndb_mgm>
mysql> select @lastepoch:=max(epoch) from mysql.ndb_apply_status;
+------------------------+
| @lastepoch:=max(epoch) |
+------------------------+
|        793593992183807 |
+------------------------+
1 row in set (0.04 sec)
根据epoch的值，找到二进制日志的位置以及文件名
mysql> select position, @firstfile:=file
-> 
-> from mysql.ndb_binlog_index
-> 
-> where epoch > @lastepoch
-> 
-> order by epoch asc
-> 
-> limit 1;
empty set (0.03 sec)
8.11 根据时间点恢复
找出恢复的时候需要用到的除第一个日志文件以外的其他的二进制日志文件
select distinct file 
from mysql.ndb_binlog_index
where epoch > @lastepoch 
and file @firstfile 
order by file;
然后进行二进制日志的恢复：
mysqlbinlog -h --set-charset=utf8 -d --start-position=829 ./mysql-bin.000012 | grep -v reload database |mysql bg
mysqlbinlog -h --set-charset=utf8 -d --stop-datetime=2012-07-18 13:30:00 ./mysql-bin.0000013 | grep -v reload database |mysql  bg
恢复完成后，退出单用户模式，并启动另外一个sql节点
ndb_mgm> exit single user mode;
exiting single user mode in progress.
use all status or show to see when single user mode has been exited.
ndb_mgm>
进入mysql节点，查到数据已经恢复了。
bitscn.com