xtrabackup
Xtrabackup 是一个对 InnoDB 做数据备份的工具，支持在线热备份(备份时不影响数据读写)
xtrabackup备份MySQL的过程:
(1)先拷贝ibdata1文件、undo文件、InnoDB/XtraDB引擎对应的表ibd等文件，不包括.frm文件
(2)flush table with read lock;
(3)拷贝MyISAM表对应的.MYD、.MYI、.frm文件,以及其他的.frm, .MRG, .MYD, .MYI, .TRG, .TRN, .ARM, .ARZ, .CSM, .CSV, .par, and .opt 文件
(4)拷贝redo log，并通过show master sataus 获得binlog位置；
(5)unlock tables;
(6)exit;
还原过程：
(1)读my.cnf文件，获得datadir,innodb_data_home_dir等目录位置
(2)先拷贝MyISAM表，索引，再拷贝InnoDB表和索引，最后是日志文件
(3)保留文件属性，改变文件所有者为备份的用户
还原时间点
还原到xtrabackup结束的时候，而不是开始备份的时候。
版本使用建议
5.5 -> 2.2
5.6 -> 2.3
5.7 -> 2.4
安装xtrabackup 2.4
sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql:>sudo yum install -y https://repo.percona.com/yum/percona-release-latest.noarch.rpm
sht-sgmhadoopnn-01:mysqladmin:/etc/yum.repos.d:>rpm -qa|grep percona
sht-sgmhadoopnn-01:mysqladmin:/etc/yum.repos.d:>sudo yum list | grep percona
sht-sgmhadoopnn-01:mysqladmin:/etc/yum.repos.d:>sudo yum install -y percona-xtrabackup-24
卸载xtrabackup
yum remove percona-xtrabackup
创建备份专用用户
mysql> CREATE USER 'bkpuser'@'%' IDENTIFIED BY 'agm43gadsg';
mysql> GRANT RELOAD, LOCK TABLES, PROCESS, REPLICATION CLIENT ON *.* TO 'bkpuser'@'%';
innobackupex中不建议使用的两个参数：
--flush-log 表示刷新日志，则在备份的时候，如果有大事物发生会造成备份失败；
--no-lock 表示不加FTWRL，则在备份的时候，如果有DDL操作会造成备份失败，或者无法还原；
xtrabackup全备+还原操作过程
(1)创建一个备份
sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql/data/mysql:> xtrabackup --user=bkpuser --password='agm43gadsg' --backup --target-dir=/usr/local/mysql/backup/full_backup_$(date +%Y%m%d%H%M%S)/ 190411 15:57:09 version_check Connected to MySQL server 190411 15:57:09 version_check Executing a version check against the server... 190411 15:57:09 version_check Done. 190411 15:57:09 Connecting to MySQL server host: localhost, user: bkpuser, password: set, port: 3306, socket: /usr/local/mysql/data/mysql.sock Using server version 5.7.21-log xtrabackup version 2.4.13 based on MySQL server 5.7.19 Linux (x86_64) (revision id: 3e7ca7c) xtrabackup: uses posix_fadvise(). xtrabackup: cd to /usr/local/mysql/data xtrabackup: open files limit requested 0, set to 1024 xtrabackup: using the following InnoDB configuration: xtrabackup: innodb_data_home_dir = /usr/local/mysql/data/ xtrabackup: innodb_data_file_path = ibdata1:200M:autoextend xtrabackup: innodb_log_group_home_dir = /usr/local/mysql/arch xtrabackup: innodb_log_files_in_group = 2 xtrabackup: innodb_log_file_size = 52428800 2019-04-11 15:57:09 0x7f9d712ba880 InnoDB: Using Linux native AIO InnoDB: Number of pools: 1 190411 15:57:09 >> log scanned up to (2388451889) InnoDB: Opened 3 undo tablespaces InnoDB: 0 undo tablespaces made active xtrabackup: Generating a list of tablespaces InnoDB: Allocated tablespace ID 75 for employees/txt_t1, old maximum was 3 190411 15:57:09 [01] Copying /usr/local/mysql/data/ibdata1 to /usr/local/mysql/backup/full_backup_20190411155708/ibdata1 190411 15:57:10 >> log scanned up to (2388451889) 190411 15:57:11 >> log scanned up to (2388451889) 190411 15:57:12 >> log scanned up to (2388451889) 190411 15:57:13 >> log scanned up to (2388451889) 190411 15:57:13 [01] ...done 190411 15:57:14 >> log scanned up to (2388451889) 190411 15:57:15 >> log scanned up to (2388451889) 190411 15:57:16 >> log scanned up to (2388451889) 190411 15:57:16 [01] Copying /usr/local/mysql/arch/undo001 to /usr/local/mysql/backup/full_backup_20190411155708/undo001 190411 15:57:16 [01] ...done 190411 15:57:17 >> log scanned up to (2388451889) 190411 15:57:17 [01] Copying /usr/local/mysql/arch/undo002 to /usr/local/mysql/backup/full_backup_20190411155708/undo002 190411 15:57:17 [01] ...done 190411 15:57:17 [01] Copying /usr/local/mysql/arch/undo003 to /usr/local/mysql/backup/full_backup_20190411155708/undo003 190411 15:57:17 [01] ...done 190411 15:57:18 [01] Copying ./employees/txt_t1.ibd to /usr/local/mysql/backup/full_backup_20190411155708/employees/txt_t1.ibd 190411 15:57:18 [01] ...done 190411 15:57:18 [01] Copying ./employees/departments.ibd to /usr/local/mysql/backup/full_backup_20190411155708/employees/departments.ibd ...... ...... 190411 15:58:28 [01] Copying ./testdb/t1.ibd to /usr/local/mysql/backup/full_backup_20190411155708/testdb/t1.ibd 190411 15:58:28 [01] ...done 190411 15:58:28 [01] Copying ./testdb/t_org_role_user.ibd to /usr/local/mysql/backup/full_backup_20190411155708/testdb/t_org_role_user.ibd 190411 15:58:28 [01] ...done 190411 15:58:29 >> log scanned up to (2388451889) 190411 15:58:29 Executing FLUSH NO_WRITE_TO_BINLOG TABLES... 190411 15:58:29 Executing FLUSH TABLES WITH READ LOCK... 190411 15:58:29 Starting to backup non-InnoDB tables and files 190411 15:58:29 [01] Copying ./employees/db.opt to /usr/local/mysql/backup/full_backup_20190411155708/employees/db.opt 190411 15:58:29 [01] ...done 190411 15:58:29 [01] Copying ./employees/txt_t1.frm to /usr/local/mysql/backup/full_backup_20190411155708/employees/txt_t1.frm 190411 15:58:33 [01] Copying ./sys/x@0024waits_by_host_by_latency.frm to /usr/local/mysql/backup/full_backup_20190411155708/sys/x@0024waits_by_host_by_latency.frm 190411 15:58:33 [01] ...done 190411 15:58:33 [01] Copying ./sys/x@0024waits_by_user_by_latency.frm to /usr/local/mysql/backup/full_backup_20190411155708/sys/x@0024waits_by_user_by_latency.frm 190411 15:58:33 [01] ...done 190411 15:58:33 [01] Copying ./sys/x@0024waits_global_by_latency.frm to /usr/local/mysql/backup/full_backup_20190411155708/sys/x@0024waits_global_by_latency.frm 190411 15:58:33 [01] ...done 190411 15:58:33 [01] Copying ./testdb/db.opt to /usr/local/mysql/backup/full_backup_20190411155708/testdb/db.opt 190411 15:58:33 [01] ...done 190411 15:58:33 [01] Copying ./testdb/geom.frm to /usr/local/mysql/backup/full_backup_20190411155708/testdb/geom.frm 190411 15:58:33 [01] ...done 190411 15:58:33 [01] Copying ./testdb/t1.frm to /usr/local/mysql/backup/full_backup_20190411155708/testdb/t1.frm 190411 15:58:33 [01] ...done 190411 15:58:33 [01] Copying ./testdb/t2.MYD to /usr/local/mysql/backup/full_backup_20190411155708/testdb/t2.MYD 190411 15:58:33 [01] ...done 190411 15:58:33 [01] Copying ./testdb/t2.MYI to /usr/local/mysql/backup/full_backup_20190411155708/testdb/t2.MYI 190411 15:58:33 [01] ...done 190411 15:58:33 [01] Copying ./testdb/t2.frm to /usr/local/mysql/backup/full_backup_20190411155708/testdb/t2.frm 190411 15:58:33 [01] ...done 190411 15:58:33 [01] Copying ./testdb/t_org_role_user.frm to /usr/local/mysql/backup/full_backup_20190411155708/testdb/t_org_role_user.frm 190411 15:58:33 [01] ...done 190411 15:58:33 Finished backing up non-InnoDB tables and files 190411 15:58:33 [00] Writing /usr/local/mysql/backup/full_backup_20190411155708/xtrabackup_binlog_info 190411 15:58:33 [00] ...done 190411 15:58:33 Executing FLUSH NO_WRITE_TO_BINLOG ENGINE LOGS... xtrabackup: The latest check point (for incremental): '2388451880' xtrabackup: Stopping log copying thread. .190411 15:58:33 >> log scanned up to (2388451889) ...... ...... 190411 15:58:33 Executing UNLOCK TABLES 190411 15:58:33 All tables unlocked 190411 15:58:33 [00] Copying ib_buffer_pool to /usr/local/mysql/backup/full_backup_20190411155708/ib_buffer_pool 190411 15:58:33 [00] ...done 190411 15:58:33 Backup created in directory '/usr/local/mysql/backup/full_backup_20190411155708/' MySQL binlog position: filename 'mysql-bin.000002', position '722', GTID of the last change '1a598743-53aa-11e9-9c46-00505682501e:1-3' 190411 15:58:33 [00] Writing /usr/local/mysql/backup/full_backup_20190411155708/backup-my.cnf 190411 15:58:33 [00] ...done 190411 15:58:33 [00] Writing /usr/local/mysql/backup/full_backup_20190411155708/xtrabackup_info 190411 15:58:33 [00] ...done xtrabackup: Transaction log of lsn (2388451880) to (2388451889) was copied. 190411 15:58:34 completed OK!
(2)准备一个备份：
原理是在还原备份之前需要做一次crash recovery，应用redolog,undolog
sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql/data/mysql:> xtrabackup --user=bkpuser --password='agm43gadsg' --prepare --target-dir=/usr/local/mysql/backup/full_backup_20190411155708/ xtrabackup: xtrabackup_logfile detected: size=8388608, start_lsn=(2388451880) xtrabackup: using the following InnoDB configuration for recovery: xtrabackup: innodb_data_home_dir = . xtrabackup: innodb_data_file_path = ibdata1:200M:autoextend xtrabackup: innodb_log_group_home_dir = . xtrabackup: innodb_log_files_in_group = 1 xtrabackup: innodb_log_file_size = 8388608 xtrabackup: using the following InnoDB configuration for recovery: xtrabackup: innodb_data_home_dir = . xtrabackup: innodb_data_file_path = ibdata1:200M:autoextend xtrabackup: innodb_log_group_home_dir = . xtrabackup: innodb_log_files_in_group = 1 xtrabackup: innodb_log_file_size = 8388608 xtrabackup: Starting InnoDB instance for recovery. xtrabackup: Using 104857600 bytes for buffer pool (set by --use-memory parameter) InnoDB: PUNCH HOLE support available InnoDB: Mutexes and rw_locks use GCC atomic builtins InnoDB: Uses event mutexes InnoDB: GCC builtin __atomic_thread_fence() is used for memory barrier InnoDB: Compressed tables use zlib 1.2.7 InnoDB: Number of pools: 1 InnoDB: Using CPU crc32 instructions InnoDB: Initializing buffer pool, total size = 100M, instances = 1, chunk size = 100M InnoDB: Completed initialization of buffer pool InnoDB: If the mysqld execution user is authorized, page cleaner thread priority can be changed. See the man page of setpriority(). InnoDB: Opened 3 undo tablespaces InnoDB: 3 undo tablespaces made active InnoDB: Highest supported file format is Barracuda. InnoDB: Log scan progressed past the checkpoint lsn 2388451880 InnoDB: Doing recovery: scanned up to log sequence number 2388451889 (0%) InnoDB: Database was not shutdown normally! InnoDB: Starting crash recovery. InnoDB: xtrabackup: Last MySQL binlog file position 722, file name mysql-bin.000002 InnoDB: Creating shared tablespace for temporary tables InnoDB: Setting file './ibtmp1' size to 12 MB. Physically writing the file full; Please wait ... InnoDB: File './ibtmp1' size is now 12 MB. InnoDB: 96 redo rollback segment(s) found. 1 redo rollback segment(s) are active. InnoDB: 32 non-redo rollback segment(s) are active. InnoDB: 5.7.19 started; log sequence number 2388451889 InnoDB: xtrabackup: Last MySQL binlog file position 722, file name mysql-bin.000002 xtrabackup: starting shutdown with innodb_fast_shutdown = 1 InnoDB: FTS optimize thread exiting. InnoDB: Starting shutdown... InnoDB: Shutdown completed; log sequence number 2388452025 InnoDB: Number of pools: 1 xtrabackup: using the following InnoDB configuration for recovery: xtrabackup: innodb_data_home_dir = . xtrabackup: innodb_data_file_path = ibdata1:200M:autoextend xtrabackup: innodb_log_group_home_dir = . xtrabackup: innodb_log_files_in_group = 2 xtrabackup: innodb_log_file_size = 52428800 InnoDB: PUNCH HOLE support available InnoDB: Mutexes and rw_locks use GCC atomic builtins InnoDB: Uses event mutexes InnoDB: GCC builtin __atomic_thread_fence() is used for memory barrier InnoDB: Compressed tables use zlib 1.2.7 InnoDB: Number of pools: 1 InnoDB: Using CPU crc32 instructions InnoDB: Initializing buffer pool, total size = 100M, instances = 1, chunk size = 100M InnoDB: Completed initialization of buffer pool InnoDB: If the mysqld execution user is authorized, page cleaner thread priority can be changed. See the man page of setpriority(). InnoDB: Setting log file ./ib_logfile101 size to 50 MB InnoDB: Setting log file ./ib_logfile1 size to 50 MB InnoDB: Renaming log file ./ib_logfile101 to ./ib_logfile0 InnoDB: New log files created, LSN=2388452025 InnoDB: Opened 3 undo tablespaces InnoDB: 3 undo tablespaces made active InnoDB: Highest supported file format is Barracuda. InnoDB: Log scan progressed past the checkpoint lsn 2388452364 InnoDB: Doing recovery: scanned up to log sequence number 2388452373 (0%) InnoDB: Database was not shutdown normally! InnoDB: Starting crash recovery. InnoDB: xtrabackup: Last MySQL binlog file position 722, file name mysql-bin.000002 InnoDB: Removed temporary tablespace data file: "ibtmp1" InnoDB: Creating shared tablespace for temporary tables InnoDB: Setting file './ibtmp1' size to 12 MB. Physically writing the file full; Please wait ... InnoDB: File './ibtmp1' size is now 12 MB. InnoDB: 96 redo rollback segment(s) found. 1 redo rollback segment(s) are active. InnoDB: 32 non-redo rollback segment(s) are active. InnoDB: page_cleaner: 1000ms intended loop took 4420ms. The settings might not be optimal. (flushed=0 and evicted=0, during the time.) InnoDB: 5.7.19 started; log sequence number 2388452373 xtrabackup: starting shutdown with innodb_fast_shutdown = 1 InnoDB: FTS optimize thread exiting. InnoDB: Starting shutdown... InnoDB: Shutdown completed; log sequence number 2388452392 190411 16:05:22 completed OK!
(3)还原一个备份
这次还原过程是在另外一台主机上，为了不破坏源主机数据，需要在还原主机上安装好相同版本的MySQL软件和配置文件my.cnf
为了保持数据库配置一致性，这里测试直接把源数据库软件拷贝到新的主机sht-sgmhadoopnn-02上：
sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql/data:>rsync -avz --progress /usr/local/mysql/* sht-sgmhadoopnn-02:/usr/local/mysql/
还原之前确保还原数据库sht-sgmhadoopnn-02 mysql shutdown，并且datadir为空，这一步可以通过rsync或cp直接拷贝备份集文件到指定data目录，也可以用--copy-back参数
sht-sgmhadoopnn-02:mysqladmin:/usr/local/mysql:>rm -rf /usr/local/mysql/data/* /usr/local/mysql/arch/* sht-sgmhadoopnn-02:mysqladmin:/usr/local/mysql:> xtrabackup --user=bkpuser --password='agm43gadsg' --host=sht-sgmhadoopnn-02 --copy-back --target-dir=/usr/local/mysql/backup/full_backup_20190411155708 xtrabackup version 2.4.13 based on MySQL server 5.7.19 Linux (x86_64) (revision id: 3e7ca7c) 190331 19:34:43 [01] Copying undo001 to /usr/local/mysql/arch/undo001 190331 19:34:43 [01] ...done 190331 19:34:43 [01] Copying undo002 to /usr/local/mysql/arch/undo002 190331 19:34:43 [01] ...done 190331 19:34:44 [01] Copying undo003 to /usr/local/mysql/arch/undo003 190331 19:34:44 [01] ...done 190331 19:34:44 [01] Copying ib_logfile0 to /usr/local/mysql/arch/ib_logfile0 190331 19:34:45 [01] ...done 190331 19:34:46 [01] Copying ib_logfile1 to /usr/local/mysql/arch/ib_logfile1 190331 19:34:47 [01] ...done 190331 19:34:48 [01] Copying ibdata1 to /usr/local/mysql/data/ibdata1 190331 19:34:52 [01] ...done 190331 19:34:55 [01] Copying ./employees/db.opt to /usr/local/mysql/data/employees/db.opt 190331 19:34:55 [01] ...done 190331 19:34:55 [01] Copying ./employees/txt_t1.frm to /usr/local/mysql/data/employees/txt_t1.frm 190331 19:34:55 [01] ...done 190331 19:34:55 [01] Copying ./employees/departments.frm to /usr/local/mysql/data/employees/departments.frm 190331 19:34:55 [01] ...done ...... ...... 190331 19:36:10 [01] Copying ./testdb/t_org_role_user.frm to /usr/local/mysql/data/testdb/t_org_role_user.frm 190331 19:36:10 [01] ...done 190331 19:36:10 [01] Copying ./testdb/t_org_role_user.ibd to /usr/local/mysql/data/testdb/t_org_role_user.ibd 190331 19:36:10 [01] ...done 190331 19:36:10 [01] Copying ./ib_buffer_pool to /usr/local/mysql/data/ib_buffer_pool 190331 19:36:10 [01] ...done 190331 19:36:10 [01] Copying ./ibtmp1 to /usr/local/mysql/data/ibtmp1 190331 19:36:10 [01] ...done 190331 19:36:11 [01] Copying ./xtrabackup_binlog_pos_innodb to /usr/local/mysql/data/xtrabackup_binlog_pos_innodb 190331 19:36:11 [01] ...done 190331 19:36:11 [01] Copying ./xtrabackup_info to /usr/local/mysql/data/xtrabackup_info 190331 19:36:11 [01] ...done 190331 19:36:11 [01] Copying ./xtrabackup_master_key_id to /usr/local/mysql/data/xtrabackup_master_key_id 190331 19:36:11 [01] ...done 190331 19:36:11 completed OK!
启动数据库
sht-sgmhadoopnn-02:mysqladmin:/usr/local/mysql:>mysqld &
[1] 6183
xtrabackup增量备份+还原操作过程
(1)先做个全备
sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql:>xtrabackup --backup --target-dir=/usr/local/mysql/backup/full_backup_$(date +%Y%m%d%H%M%S) sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql:>ll /usr/local/mysql/backup/ drwxr-x--- 7 mysqladmin dba 299 Apr 11 16:50 full_backup_20190411164929 插入测试数据： mysql> create table testdb.t3(c1 int,c2 varchar(20)); mysql> insert into t1 values(1,'aaa');
(2)增量备份1+增量备份2
sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql:> xtrabackup --backup --target-dir=/usr/local/mysql/backup/inc1_backup_$(date +%Y%m%d%H%M%S) --incremental-basedir=/usr/local/mysql/backup/full_backup_20190411164929 sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql:>ll backup/ total 4 drwxr-x--- 7 mysqladmin dba 299 Apr 11 16:50 full_backup_20190411164929 drwxr-x--- 7 mysqladmin dba 4096 Apr 11 17:02 inc1_backup_20190411170239 mysql> insert into t1 values(2,'bbb'); sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql:> xtrabackup --backup --target-dir=/usr/local/mysql/backup/inc2_backup_$(date +%Y%m%d%H%M%S) --incremental-basedir=/usr/local/mysql/backup/inc1_backup_20190411170239 sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql:>ll backup/ total 8 drwxr-x--- 7 mysqladmin dba 299 Apr 11 16:50 full_backup_20190411164929 drwxr-x--- 7 mysqladmin dba 4096 Apr 11 17:02 inc1_backup_20190411170239 drwxr-x--- 7 mysqladmin dba 4096 Apr 12 15:58 inc2_backup_20190412155845
(3)准备一个增量备份
sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql:> xtrabackup --prepare --apply-log-only --target-dir=/usr/local/mysql/backup/full_backup_20190411164929 sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql:> xtrabackup --prepare --apply-log-only --target-dir=/usr/local/mysql/backup/full_backup_20190411164929 --incremental-dir=/usr/local/mysql/backup/inc1_backup_20190411170239 sht-sgmhadoopnn-01:mysqladmin:/usr/local/mysql:> xtrabackup --prepare --target-dir=/usr/local/mysql/backup/full_backup_20190411164929 --incremental-dir=/usr/local/mysql/backup/inc2_backup_20190412155845
(4)还原一个备份
这次还原过程是在另外一台主机上，为了不破坏源主机数据，需要在还原主机上安装好相同版本的MySQL软件和配置文件my.cnf
为了保持数据库配置一致性，这里测试直接把备份文件贝到新的主机sht-sgmhadoopnn-02上：
sht-sgmhadoopnn-02:mysqladmin:/usr/local/mysql:> rm -rf /usr/local/mysql/data/* /usr/local/mysql/arch/* sht-sgmhadoopnn-02:mysqladmin:/usr/local/mysql:> rsync -avz --progress /usr/local/mysql/backup/full_backup_20190411164929 sht-sgmhadoopnn-02:/usr/local/mysql/backup/ sht-sgmhadoopnn-02:mysqladmin:/usr/local/mysql:> xtrabackup --copy-back --target-dir=/usr/local/mysql/backup/full_backup_20190411164929 sht-sgmhadoopnn-02:mysqladmin:/usr/local/mysql:>mysqld & 检查数据： mysql> select * from t1; +------+------+ | c1 | c2 | +------+------+ | 1 | aaa | | 2 | bbb | +------+------+ 12 rows in set (0.00 sec)
xtrabackup备份可能遇到的问题:
MySQL数据库只有50GB，测试的时候只要半小时就能备份完成，但是突然一次xtrabackup运行从1点到6点才完成
查看备份日志，发现从执行flush table with read lock，到unlock用了5个多小时，这个过程是copy一些non-InnoDB文件，正常情况非常快，但是这里却用了这么久时间
可能的原因：
执行FTWRL之前有慢select或耗时的DDL，DML操作，堵塞了FTWRL操作，然后再向数据库做一些DDL,DML都要等待FTWRL动作完成并unlock;
在三个不同的会话上依次执行下面三条SQL
session1:一条慢SQL mysql> select count(*) from employees e1 join dept_emp; session2:被s1阻塞 mysql> flush table with read lock; session3:被s2阻塞 mysql> insert into t1 values(1,'ccc'); mysql> show full processlist; +----+------+-----------+-----------+---------+------+------------------------------+-------------------------------------------------+ | Id | User | Host | db | Command | Time | State | Info | +----+------+-----------+-----------+---------+------+------------------------------+-------------------------------------------------+ | 39 | root | localhost | employees | Query | 48 | Sending data | select count(*) from employees e1 join dept_emp | | 40 | root | localhost | employees | Query | 32 | Waiting for table flush | flush table with read lock | | 41 | root | localhost | employees | Query | 21 | Waiting for global read lock | insert into t1 values(1,'ccc') | | 42 | root | localhost | NULL | Query | 0 | starting | show full processlist | +----+------+-----------+-----------+---------+------+------------------------------+-------------------------------------------------+ 4 rows in set (0.00 sec)
解决方法：
在业务不繁忙的时候做xtrabackup;
在从库上面备份；
指定参数--ftwrl-wait-timeout=10 #FTWRL等待超过10s，则备份报错，自动退出
xtrabackup --backup --target-dir=/usr/local/mysql/backup/full_backup_$(date +%Y%m%d%H%M%S) --ftwrl-wait-timeout=10 190413 19:36:26 Waiting for query 56 (duration 132 sec): select count(*) from employees e1 join dept_emp190413 19:36:27 >> log scanned up to (2388459962) 190413 19:36:27 Waiting for query 56 (duration 133 sec): select count(*) from employees e1 join dept_emp190413 19:36:28 >> log scanned up to (2388459962) 190413 19:36:28 Waiting for query 56 (duration 134 sec): select count(*) from employees e1 join dept_emp190413 19:36:29 >> log scanned up to (2388459962) 190413 19:36:29 Unable to obtain lock. Please try again later.
--kill-long-queries-timeout=10 #FTWRL等待超过10s后，kill掉慢SQL,备份完成
mysql> select count(*) from employees e1 join dept_emp; xtrabackup --backup --target-dir=/usr/local/mysql/backup/full_backup_$(date +%Y%m%d%H%M%S) --kill-long-queries-timeout=10 ...... 190413 19:31:01 Executing FLUSH TABLES WITH READ LOCK... 190413 19:31:01 Kill query timeout 10 seconds. 190413 19:31:11 Connecting to MySQL server host: localhost, user: root, password: set, port: 3306, socket: /usr/local/mysql/data/mysql.sock 190413 19:31:11 Killing query 39 (duration 980 sec): select count(*) from employees e1 join dept_emp 190413 19:31:11 Killing query 39 (duration 980 sec): select count(*) from employees e1 join dept_emp