前言
在發生故障切換後,經常遇到的問題就是同步報錯,數據庫很小的時候,dump完再導入很簡單就處理好了,但線上的數據庫都150G-200G,如果用單純的這種方法,成本太高,故經過一段時間的摸索,總結了幾種處理方法。
生產環境架構圖
目前現網的架構,保存著兩份數據,通過異步復制做的高可用集群,兩台機器提供對外服務。在發生故障時,切換到slave上,並將其變成master,壞掉的機器反向同步新的master,在處理故障時,遇到最多的就是主從報錯。下面是我收錄下來的報錯信息。
常見錯誤
最常見的3種情況
這3種情況是在HA切換時,由於是異步復制,且sync_binlog=0,會造成一小部分binlog沒接收完導致同步報錯。
第一種:在master上刪除一條記錄,而slave上找不到。
Last_SQL_Error: Could not execute Delete_rows event on table hcy.t1;
Can't find record in 't1',
Error_code: 1032; handler error HA_ERR_KEY_NOT_FOUND;
the event's master log mysql-bin.000006, end_log_pos 254
第二種:主鍵重復。在slave已經有該記錄,又在master上插入了同一條記錄。
Last_SQL_Error: Could not execute Write_rows event on table hcy.t1;
Duplicate entry '2' for key 'PRIMARY',
Error_code: 1062;
handler error HA_ERR_FOUND_DUPP_KEY; the event's master log mysql-bin.000006, end_log_pos 924
第三種:在master上更新一條記錄,而slave上找不到,丟失了數據。
Last_SQL_Error: Could not execute Update_rows event on table hcy.t1;
Can't find record in 't1',
Error_code: 1032;
handler error HA_ERR_KEY_NOT_FOUND; the event's master log mysql-bin.000010, end_log_pos 263
異步半同步區別
異步復制
簡單的說就是master把binlog發送過去,不管slave是否接收完,也不管是否執行完,這一動作就結束了.
半同步復制
簡單的說就是master把binlog發送過去,slave確認接收完,但不管它是否執行完,給master一個信號我這邊收到了,這一動作就結束了。(谷歌寫的代碼,5.5上正式應用。)
異步的劣勢
當master上寫操作繁忙時,當前POS點例如是10,而slave上IO_THREAD線程接收過來的是3,此時master宕機,會造成相差7個點未傳送到slave上而數據丟失。
特殊的情況
slave的中繼日志relay-bin損壞。
Last_SQL_Error: Error initializing relay log position: I/O error reading the header from the binary log
Last_SQL_Error: Error initializing relay log position: Binlog has bad magic number;
It's not a binary log file that can be used by this version of MySQL
這種情況SLAVE在宕機,或者非法關機,例如電源故障、主板燒了等,造成中繼日志損壞,同步停掉。
人為失誤需謹慎:多台slave存在重復server-id
這種情況同步會一直延時,永遠也同步不完,error錯誤日志裡一直出現上面兩行信息。解決方法就是把server-id改成不一致即可。
Slave: received end packet from server, apparent master shutdown:
Slave I/O thread: Failed reading log event, reconnecting to retry, log 'mysql-bin.000012' at postion 106
問題處理
刪除失敗
在master上刪除一條記錄,而slave上找不到。
Last_SQL_Error: Could not execute Delete_rows event on table hcy.t1;
Can't find record in 't1',
Error_code: 1032; handler error HA_ERR_KEY_NOT_FOUND;
the event's master log mysql-bin.000006, end_log_pos 254
解決方法:
由於master要刪除一條記錄,而slave上找不到故報錯,這種情況主上都將其刪除了,那麼從機可以直接跳過。可用命令:
stop slave;
set global sql_slave_skip_counter=1;
start slave;
如果這種情況很多,可用我寫的一個腳本skip_error_replcation.sh,默認跳過10個錯誤(只針對這種情況才跳,其他情況輸出錯誤結果,等待處理),這個腳本是參考maakit工具包的mk-slave-restart原理用shell寫的,功能上定義了一些自己的東西,不是無論什麼錯誤都一律跳過。)
主鍵重復
在slave已經有該記錄,又在master上插入了同一條記錄。
Last_SQL_Error: Could not execute Write_rows event on table hcy.t1; Duplicate entry '2' for key 'PRIMARY', Error_code: 1062; handler error HA_ERR_FOUND_DUPP_KEY; the event's master log mysql-bin.000006, end_log_pos 924
解決方法:
在slave上用desc hcy.t1; 先看下表結構:
mysql> desc hcy.t1; +-------+---------+------+-----+---------+-------+ | Field | Type | Null | Key | Default | Extra | +-------+---------+------+-----+---------+-------+ | id | int(11) | NO | PRI | 0 | | | name | char(4) | YES | | NULL | | +-------+---------+------+-----+---------+-------+
刪除重復的主鍵
mysql> delete from t1 where id=2; Query OK, 1 row affected (0.00 sec) mysql> start slave; Query OK, 0 rows affected (0.00 sec) mysql> show slave status\G; …… Slave_IO_Running: Yes Slave_SQL_Running: Yes …… mysql> select * from t1 where id=2;
在master上和slave上再分別確認一下。
更新丟失
在master上更新一條記錄,而slave上找不到,丟失了數據。
Last_SQL_Error: Could not execute Update_rows event on table hcy.t1; Can't find record in 't1', Error_code: 1032; handler error HA_ERR_KEY_NOT_FOUND; the event's master log mysql-bin.000010, end_log_pos 794
解決方法:
在master上,用mysqlbinlog 分析下出錯的binlog日志在干什麼。
/usr/local/mysql/bin/mysqlbinlog --no-defaults -v -v --base64-output=DECODE-ROWS mysql-bin.000010 | grep -A '10' 794 #120302 12:08:36 server id 22 end_log_pos 794 Update_rows: table id 33 flags: STMT_END_F ### UPDATE hcy.t1 ### WHERE ### @1=2 /* INT meta=0 nullable=0 is_null=0 */ ### @2='bbc' /* STRING(4) meta=65028 nullable=1 is_null=0 */ ### SET ### @1=2 /* INT meta=0 nullable=0 is_null=0 */ ### @2='BTV' /* STRING(4) meta=65028 nullable=1 is_null=0 */ # at 794 #120302 12:08:36 server id 22 end_log_pos 821 Xid = 60 COMMIT/*!*/; DELIMITER ; # End of log file ROLLBACK /* added by mysqlbinlog */; /*!50003 SET COMPLETION_TYPE=@OLD_COMPLETION_TYPE*/;
在slave上,查找下更新後的那條記錄,應該是不存在的。
mysql> select * from t1 where id=2;
Empty set (0.00 sec)
然後再到master查看
mysql> select * from t1 where id=2; +----+------+ | id | name | +----+------+ | 2 | BTV | +----+------+ 1 row in set (0.00 sec)
把丟失的數據在slave上填補,然後跳過報錯即可。
mysql> insert into t1 values (2,'BTV'); Query OK, 1 row affected (0.00 sec) mysql> select * from t1 where id=2; +----+------+ | id | name | +----+------+ | 2 | BTV | +----+------+ 1 row in set (0.00 sec) mysql> stop slave ;set global sql_slave_skip_counter=1;start slave; Query OK, 0 rows affected (0.01 sec) Query OK, 0 rows affected (0.00 sec) Query OK, 0 rows affected (0.00 sec) mysql> show slave status\G; …… Slave_IO_Running: Yes Slave_SQL_Running: Yes ……
中繼日志損壞
slave的中繼日志relay-bin損壞。
Last_SQL_Error: Error initializing relay log position: I/O error reading the header from the binary log Last_SQL_Error: Error initializing relay log position: Binlog has bad magic number; It's not a binary log file that can be used by this version of MySQL
手工修復
解決方法:找到同步的binlog和POS點,然後重新做同步,這樣就可以有新的中繼日值了。
例子:
mysql> show slave status\G; *************************** 1. row *************************** Master_Log_File: mysql-bin.000010 Read_Master_Log_Pos: 1191 Relay_Log_File: vm02-relay-bin.000005 Relay_Log_Pos: 253 Relay_Master_Log_File: mysql-bin.000010 Slave_IO_Running: Yes Slave_SQL_Running: No Replicate_Do_DB: Replicate_Ignore_DB: Replicate_Do_Table: Replicate_Ignore_Table: Replicate_Wild_Do_Table: Replicate_Wild_Ignore_Table: Last_Errno: 1593 Last_Error: Error initializing relay log position: I/O error reading the header from the binary log Skip_Counter: 1 Exec_Master_Log_Pos: 821
Slave_IO_Running :接收master的binlog信息
Master_Log_File
Read_Master_Log_Pos
Slave_SQL_Running:執行寫操作
Relay_Master_Log_File
Exec_Master_Log_Pos
以執行寫的binlog和POS點為准。
Relay_Master_Log_File: mysql-bin.000010 Exec_Master_Log_Pos: 821 mysql> stop slave; Query OK, 0 rows affected (0.01 sec) mysql> CHANGE MASTER TO MASTER_LOG_FILE='mysql-bin.000010',MASTER_LOG_POS=821; Query OK, 0 rows affected (0.01 sec) mysql> start slave; Query OK, 0 rows affected (0.00 sec) mysql> show slave status\G; *************************** 1. row *************************** Slave_IO_State: Waiting for master to send event Master_Host: 192.168.8.22 Master_User: repl Master_Port: 3306 Connect_Retry: 10 Master_Log_File: mysql-bin.000010 Read_Master_Log_Pos: 1191 Relay_Log_File: vm02-relay-bin.000002 Relay_Log_Pos: 623 Relay_Master_Log_File: mysql-bin.000010 Slave_IO_Running: Yes Slave_SQL_Running: Yes Replicate_Do_DB: Replicate_Ignore_DB: Replicate_Do_Table: Replicate_Ignore_Table: Replicate_Wild_Do_Table: Replicate_Wild_Ignore_Table: Last_Errno: 0 Last_Error: Skip_Counter: 0 Exec_Master_Log_Pos: 1191 Relay_Log_Space: 778 Until_Condition: None Until_Log_File: Until_Log_Pos: 0 Master_SSL_Allowed: No Master_SSL_CA_File: Master_SSL_CA_Path: Master_SSL_Cert: Master_SSL_Cipher: Master_SSL_Key: Seconds_Behind_Master: 0 Master_SSL_Verify_Server_Cert: No Last_IO_Errno: 0 Last_IO_Error: Last_SQL_Errno: 0 Last_SQL_Error: Ibbackup
各種大招都用上了,無奈slave數據丟失過多,ibbackup(需要銀子)該你登場了。
Ibbackup熱備份工具,是付費的。xtrabackup是免費的,功能上一樣。
Ibbackup備份期間不鎖表,備份時開啟一個事務(相當於做一個快照),然後會記錄一個點,之後數據的更改保存在ibbackup_logfile文件裡,恢復時把ibbackup_logfile 變化的數據再寫入到ibdata裡。
Ibbackup 只備份數據( ibdata、.ibd ),表結構.frm不備份。
下面一個演示例子:
備份:ibbackup /bak/etc/my_local.cnf /bak/etc/my_bak.cnf
恢復:ibbackup --apply-log /bak/etc/my_bak.cnf
[root@vm01 etc]# more my_local.cnf datadir =/usr/local/mysql/data innodb_data_home_dir = /usr/local/mysql/data innodb_data_file_path = ibdata1:10M:autoextend innodb_log_group_home_dir = /usr/local/mysql/data innodb_buffer_pool_size = 100M innodb_log_file_size = 5M innodb_log_files_in_group=2 [root@vm01 etc]# ibbackup /bak/etc/my_local.cnf /bak/etc/my_bak.cnf InnoDB Hot Backup version 3.0.0; Copyright 2002-2005 Innobase Oy License A21488 is granted to vm01 ([email protected]) (--apply-log works in any computer regardless of the hostname) Licensed for use in a computer whose hostname is 'vm01' Expires 2012-5-1 (year-month-day) at 00:00 See http://www.innodb.com for further information Type ibbackup --license for detailed license terms, --help for help Contents of /bak/etc/my_local.cnf: innodb_data_home_dir got value /usr/local/mysql/data innodb_data_file_path got value ibdata1:10M:autoextend datadir got value /usr/local/mysql/data innodb_log_group_home_dir got value /usr/local/mysql/data innodb_log_files_in_group got value 2 innodb_log_file_size got value 5242880 Contents of /bak/etc/my_bak.cnf: innodb_data_home_dir got value /bak/data innodb_data_file_path got value ibdata1:10M:autoextend datadir got value /bak/data innodb_log_group_home_dir got value /bak/data innodb_log_files_in_group got value 2 innodb_log_file_size got value 5242880 ibbackup: Found checkpoint at lsn 0 1636898 ibbackup: Starting log scan from lsn 0 1636864 120302 16:47:43 ibbackup: Copying log... 120302 16:47:43 ibbackup: Log copied, lsn 0 1636898 ibbackup: We wait 1 second before starting copying the data files... 120302 16:47:44 ibbackup: Copying /usr/local/mysql/data/ibdata1 ibbackup: A copied database page was modified at 0 1636898 ibbackup: Scanned log up to lsn 0 1636898 ibbackup: Was able to parse the log up to lsn 0 1636898 ibbackup: Maximum page number for a log record 0 120302 16:47:46 ibbackup: Full backup completed! [root@vm01 etc]# [root@vm01 etc]# cd /bak/data/ [root@vm01 data]# ls ibbackup_logfile ibdata1 [root@vm01 data]# ibbackup --apply-log /bak/etc/my_bak.cnf InnoDB Hot Backup version 3.0.0; Copyright 2002-2005 Innobase Oy License A21488 is granted to vm01 ([email protected]) (--apply-log works in any computer regardless of the hostname) Licensed for use in a computer whose hostname is 'vm01' Expires 2012-5-1 (year-month-day) at 00:00 See http://www.innodb.com for further information Type ibbackup --license for detailed license terms, --help for help Contents of /bak/etc/my_bak.cnf: innodb_data_home_dir got value /bak/data innodb_data_file_path got value ibdata1:10M:autoextend datadir got value /bak/data innodb_log_group_home_dir got value /bak/data innodb_log_files_in_group got value 2 innodb_log_file_size got value 5242880 120302 16:48:38 ibbackup: ibbackup_logfile's creation parameters: ibbackup: start lsn 0 1636864, end lsn 0 1636898, ibbackup: start checkpoint 0 1636898 ibbackup: start checkpoint 0 1636898 InnoDB: Doing recovery: scanned up to log sequence number 0 1636898 InnoDB: Starting an apply batch of log records to the database... InnoDB: Progress in percents: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 .....99 Setting log file size to 0 5242880 ibbackup: We were able to parse ibbackup_logfile up to ibbackup: lsn 0 1636898 ibbackup: Last MySQL binlog file position 0 1191, file name ./mysql-bin.000010 ibbackup: The first data file is '/bak/data/ibdata1' ibbackup: and the new created log files are at '/bak/data/' 120302 16:48:38 ibbackup: Full backup prepared for recovery successfully! [root@vm01 data]# ls ibbackup_logfile ibdata1 ib_logfile0 ib_logfile1
把ibdata1 ib_logfile0 ib_logfile1拷貝到從,把.frm也拷貝過去,啟動MySQL後,做同步,那個點就是上面輸出的:
ibbackup: Last MySQL binlog file position 0 1191, file name ./mysql-bin.000010
CHANGE MASTER TO MASTER_LOG_FILE='mysql-bin.000010',MASTER_LOG_POS=1191;
Maatkit工具包
http://www.maatkit.org/
簡介
maatkit是一個開源的工具包,為mysql日常管理提供了幫助。目前,已被Percona公司收購並維護。其中:
mk-table-checksum是用來檢測master和slave上的表結構和數據是否一致。
mk-table-sync是發生主從數據不一致時,來修復的。
這兩個工具包,沒有在現網實際操作的經驗,這裡僅僅是新技術探討和學術交流,下面展示下如何使用。
http://www.actionsky.com/products/mysql-others/maatkit.jsp
[root@vm02]# mk-table-checksum h=vm01,u=admin,p=123456 h=vm02,u=admin,p=123456 -d hcy -t t1 Cannot connect to MySQL because the Perl DBI module is not installed or not found. Run 'perl -MDBI' to see the directories that Perl searches for DBI. If DBI is not installed, try: Debian/Ubuntu apt-get install libdbi-perl RHEL/CentOS yum install perl-DBI OpenSolaris pgk install pkg:/SUNWpmdbi
提示缺少perl-DBI模塊,那麼直接 yum install perl-DBI。
[root@vm02 bin]# mk-table-checksum h=vm01,u=admin,p=123456 h=vm02,u=admin,p=123456 -d hcy -t t1 DATABASE TABLE CHUNK HOST ENGINE COUNT CHECKSUM TIME WAIT STAT LAG hcy t1 0 vm02 InnoDB NULL 1957752020 0 0 NULL NULL hcy t1 0 vm01 InnoDB NULL 1957752020 0 0 NULL NULL
如果表數據不一致,CHECKSUM的值是不相等的。
解釋下輸出的意思:
DATABASE:數據庫名
TABLE:表名
CHUNK:checksum時的近似數值
HOST:MYSQL的地址
ENGINE:表引擎
COUNT:表的行數
CHECKSUM:校驗值
TIME:所用時間
WAIT:等待時間
STAT:MASTER_POS_WAIT()返回值
LAG:slave的延時時間
如果你想過濾出不相等的都有哪些表,可以用mk-checksum-filter這個工具,只要在後面加個管道符就行了。
[root@vm02 ~]# mk-table-checksum h=vm01,u=admin,p=123456 h=vm02,u=admin,p=123456 -d hcy | mk-checksum-filter hcy t2 0 vm01 InnoDB NULL 1957752020 0 0 NULL NULL hcy t2 0 vm02 InnoDB NULL 1068689114 0 0 NULL NULL
知道有哪些表不一致,可以用mk-table-sync這個工具來處理。
注:在執行mk-table-checksum時會鎖表,表的大小取決於執行的快慢。
MASTER上的t2表數據:
SLAVE上的t2表數據:
mysql> select * from t2; mysql> select * from t2; +----+------+ +----+------+ | id | name | | id | name | +----+------+ +----+------+ | 1 | a | | 1 | a | | 2 | b | | 2 | b | | 3 | ss | | 3 | ss | | 4 | asd | | 4 | asd | | 5 | ss | +----+------+ +----+------+ 4 rows in set (0.00 sec) 5 rows in set (0.00 sec) mysql> \! hostname; mysql> \! hostname; vm02 vm01 [root@vm02 ~]# mk-table-sync --execute --print --no-check-slave --transaction --databases hcy h=vm01,u=admin,p=123456 h=vm02,u=admin,p=123456 INSERT INTO `hcy`.`t2`(`id`, `name`) VALUES ('5', 'ss') /*maatkit src_db:hcy src_tbl:t2 src_dsn:h=vm01,p=...,u=admin dst_db:hcy dst_tbl:t2 dst_dsn:h=vm02,p=...,u=admin lock:0 transaction:1 changing_src:0 replicate:0 bidirectional:0 pid:3246 user:root host:vm02*/;
它的工作原理是:先一行一行檢查主從庫的表是否一樣,如果哪裡不一樣,就執行刪除,更新,插入等操作,使其達到一致。表的大小決定著執行的快慢。
If C<--transaction> is specified, C<LOCK TABLES> is not used. Instead, lock and unlock are implemented by beginning and committing transactions. The exception is if L<"--lock"> is 3. If C<--no-transaction> is specified, then C<LOCK TABLES> is used for any value of L<"--lock">. See L<"--[no]transaction">. When enabled, either explicitly or implicitly, the transaction isolation level is set C<REPEATABLE READ> and transactions are started C<WITH CONSISTENT SNAPSHOT>
MySQL復制監控
MySQL常見錯誤類型
1005:創建表失敗
1006:創建數據庫失敗
1007:數據庫已存在,創建數據庫失敗
1008:數據庫不存在,刪除數據庫失敗
1009:不能刪除數據庫文件導致刪除數據庫失敗
1010:不能刪除數據目錄導致刪除數據庫失敗
1011:刪除數據庫文件失敗
1012:不能讀取系統表中的記錄
1020:記錄已被其他用戶修改
1021:硬盤剩余空間不足,請加大硬盤可用空間
1022:關鍵字重復,更改記錄失敗
1023:關閉時發生錯誤
1024:讀文件錯誤
1025:更改名字時發生錯誤
1026:寫文件錯誤
1032:記錄不存在
1036:數據表是只讀的,不能對它進行修改
1037:系統內存不足,請重啟數據庫或重啟服務器
1038:用於排序的內存不足,請增大排序緩沖區
1040:已到達數據庫的最大連接數,請加大數據庫可用連接數
1041:系統內存不足
1042:無效的主機名
1043:無效連接
1044:當前用戶沒有訪問數據庫的權限
1045:不能連接數據庫,用戶名或密碼錯誤
1048:字段不能為空
1049:數據庫不存在
1050:數據表已存在
1051:數據表不存在
1054:字段不存在
1065:無效的SQL語句,SQL語句為空
1081:不能建立Socket連接
1114:數據表已滿,不能容納任何記錄
1116:打開的數據表太多
1129:數據庫出現異常,請重啟數據庫
1130:連接數據庫失敗,沒有連接數據庫的權限
1133:數據庫用戶不存在
1141:當前用戶無權訪問數據庫
1142:當前用戶無權訪問數據表
1143:當前用戶無權訪問數據表中的字段
1146:數據表不存在
1147:未定義用戶對數據表的訪問權限
1149:SQL語句語法錯誤
1158:網絡錯誤,出現讀錯誤,請檢查網絡連接狀況
1159:網絡錯誤,讀超時,請檢查網絡連接狀況
1160:網絡錯誤,出現寫錯誤,請檢查網絡連接狀況
1161:網絡錯誤,寫超時,請檢查網絡連接狀況
1062:字段值重復,入庫失敗
1169:字段值重復,更新記錄失敗
1177:打開數據表失敗
1180:提交事務失敗
1181:回滾事務失敗
1203:當前用戶和數據庫建立的連接已到達數據庫的最大連接數,請增大可用的數據庫連接數或重啟數據庫
1205:加鎖超時
1211:當前用戶沒有創建用戶的權限
1216:外鍵約束檢查失敗,更新子表記錄失敗
1217:外鍵約束檢查失敗,刪除或修改主表記錄失敗
1226:當前用戶使用的資源已超過所允許的資源,請重啟數據庫或重啟服務器
1227:權限不足,您無權進行此操作
1235:MySQL版本過低,不具有本功能
復制監控腳本
參考原文修改。
原腳本
#!/bin/bash # #check_mysql_slave_replication_status # # # parasum=2 help_msg(){ cat << help +---------------------+ +Error Cause: +you must input $parasum parameters! +1st : Host_IP +2st : Host_Port help exit } [ $# -ne ${parasum} ] && help_msg #若參數不夠打印幫助信息並退出 export HOST_IP=$1 export HOST_PORt=$2 MYUSER="root" MYPASS="123456" MYSQL_CMD="mysql -u$MYUSER -p$MYPASS" MailTitle="" #郵件主題 Mail_Address_MysqlStatus="[email protected]" #收件人郵箱 time1=$(date +"%Y%m%d%H%M%S") time2=$(date +"%Y-%m-%d %H:%M:%S") SlaveStatusFile=/tmp/salve_status_${HOST_PORT}.${time1} #郵件內容所在文件 echo "--------------------Begin at: "$time2 > $SlaveStatusFile echo "" >> $SlaveStatusFile #get slave status ${MYSQL_CMD} -e "show slave status\G" >> $SlaveStatusFile #取得salve進程的狀態 #get io_thread_status,sql_thread_status,last_errno 取得以下狀態值 IOStatus=$(cat $SlaveStatusFile|grep Slave_IO_Running|awk '{print $2}') SQLStatus=$(cat $SlaveStatusFile|grep Slave_SQL_Running |awk '{print $2}') Errno=$(cat $SlaveStatusFile|grep Last_Errno | awk '{print $2}') Behind=$(cat $SlaveStatusFile|grep Seconds_Behind_Master | awk '{print $2}') echo "" >> $SlaveStatusFile if [ "$IOStatus" == "No" ] || [ "$SQLStatus" == "No" ];then #判斷錯誤類型 if [ "$Errno" -eq 0 ];then #可能是salve線程未啟動 $MYSQL_CMD -e "start slave io_thread;start slave sql_thread;" echo "Cause slave threads doesnot's running,trying start slsave io_thread;start slave sql_thread;" >> $SlaveStatusFile MailTitle="[Warning] Slave threads stoped on $HOST_IP $HOST_PORT" elif [ "$Errno" -eq 1007 ] || [ "$Errno" -eq 1053 ] || [ "$Errno" -eq 1062 ] || [ "$Errno" -eq 1213 ] || [ "$Errno" -eq 1032 ]\ || [ "Errno" -eq 1158 ] || [ "$Errno" -eq 1159 ] || [ "$Errno" -eq 1008 ];then #忽略此些錯誤 $MYSQL_CMD -e "stop slave;set global sql_slave_skip_counter=1;start slave;" echo "Cause slave replication catch errors,trying skip counter and restart slave;stop slave ;set global sql_slave_skip_counter=1;slave start;" >> $SlaveStatusFile MailTitle="[Warning] Slave error on $HOST_IP $HOST_PORT! ErrNum: $Errno" else echo "Slave $HOST_IP $HOST_PORT is down!" >> $SlaveStatusFile MailTitle="[ERROR]Slave replication is down on $HOST_IP $HOST_PORT ! ErrNum:$Errno" fi fi if [ -n "$Behind" ];then Behind=0 fi echo "$Behind" >> $SlaveStatusFile #delay behind master 判斷延時時間 if [ $Behind -gt 300 ];then echo `date +"%Y-%m%d %H:%M:%S"` "slave is behind master $Bebind seconds!" >> $SlaveStatusFile MailTitle="[Warning]Slave delay $Behind seconds,from $HOST_IP $HOST_PORT" fi if [ -n "$MailTitle" ];then #若出錯或者延時時間大於300s則發送郵件 cat ${SlaveStatusFile} | /bin/mail -s "$MailTitle" $Mail_Address_MysqlStatus fi #del tmpfile:SlaveStatusFile > $SlaveStatusFile
修改後腳本
只做了簡單的整理,修正了Behind為NULL的判斷,但均未測試;
應可考慮增加:
對修復執行結果的判斷;多條錯誤的循環修復、檢測、再修復?
取消SlaveStatusFile臨時文件。
Errno、Behind兩種告警分別發郵件,告警正文增加show slave結果原文。
增加PATH,以便加到crontab中。
考慮crontab中周期執行(加鎖避免執行沖突、執行周期選擇)
增加執行日志?
#!/bin/sh # check_mysql_slave_replication_status # 參考:http://www.tianfeiyu.com/?p=2062 Usage(){ echo Usage: echo "$0 HOST PORT USER PASS" } [ -z "$1" -o -z "$2" -o -z "$3" -o -z "$4" ] && Usage && exit 1 HOST=$1 PORT=$2 USER=$3 PASS=$4 MYSQL_CMD="mysql -h$HOST -P$PORT -u$USER -p$PASS" MailTitle="" #郵件主題 Mail_Address_MysqlStatus="[email protected]" #收件人郵箱 time1=$(date +"%Y%m%d%H%M%S") time2=$(date +"%Y-%m-%d %H:%M:%S") SlaveStatusFile=/tmp/salve_status_${HOST_PORT}.${time1} #郵件內容所在文件 echo "--------------------Begin at: "$time2 > $SlaveStatusFile echo "" >> $SlaveStatusFile #get slave status ${MYSQL_CMD} -e "show slave status\G" >> $SlaveStatusFile #取得salve進程的狀態 #get io_thread_status,sql_thread_status,last_errno 取得以下狀態值 IOStatus=$(cat $SlaveStatusFile|grep Slave_IO_Running|awk '{print $2}') SQLStatus=$(cat $SlaveStatusFile|grep Slave_SQL_Running |awk '{print $2}') Errno=$(cat $SlaveStatusFile|grep Last_Errno | awk '{print $2}') Behind=$(cat $SlaveStatusFile|grep Seconds_Behind_Master | awk '{print $2}') echo "" >> $SlaveStatusFile if [ "$IOStatus" = "No" -o "$SQLStatus" = "No" ];then case "$Errno" in 0) # 可能是slave未啟動 $MYSQL_CMD -e "start slave io_thread;start slave sql_thread;" echo "Cause slave threads doesnot's running,trying start slsave io_thread;start slave sql_thread;" >> $SlaveStatusFile ;; 1007|1053|1062|1213|1032|1158|1159|1008) # 忽略這些錯誤 $MYSQL_CMD -e "stop slave;set global sql_slave_skip_counter=1;start slave;" echo "Cause slave replication catch errors,trying skip counter and restart slave;stop slave ;set global sql_slave_skip_counter=1;slave start;" >> $SlaveStatusFile MailTitle="[Warning] Slave error on $HOST:$PORT! ErrNum: $Errno" ;; *) echo "Slave $HOST:$PORT is down!" >> $SlaveStatusFile MailTitle="[ERROR]Slave replication is down on $HOST:$PORT! Errno:$Errno" ;; esac fi if [ "$Behind" = "NULL" -o -z "$Behind" ];then Behind=0 fi echo "Behind:$Behind" >> $SlaveStatusFile #delay behind master 判斷延時時間 if [ $Behind -gt 300 ];then echo `date +"%Y-%m%d %H:%M:%S"` "slave is behind master $Bebind seconds!" >> $SlaveStatusFile MailTitle="[Warning]Slave delay $Behind seconds,from $HOST $PORT" fi if [ -n "$MailTitle" ];then #若出錯或者延時時間大於300s則發送郵件 cat ${SlaveStatusFile} | /bin/mail -s "$MailTitle" $Mail_Address_MysqlStatus fi #del tmpfile:SlaveStatusFile > $SlaveStatusFile
以上這篇線上MYSQL同步報錯故障處理方法總結(必看篇)就是小編分享給大家的全部內容了,希望能給大家一個參考,也希望大家多多支持。