使用innodb_force_recovery解決MySQL崩潰無法重啟問題

更新時間：2015年05月29日 10:40:02 投稿：junjie

這篇文章主要介紹了使用innodb_force_recovery解決MySQL崩潰無法重啟問題,這只一個成功案例,并不是萬能的解決方法,需要酌情考慮,需要的朋友可以參考下

一背景

某一創(chuàng)業(yè)的朋友的主機因為磁盤陣列損壞機器crash,重啟MySQL服務時報如下錯誤:

InnoDB: Reading tablespace information from the .ibd files...

InnoDB: Restoring possible half-written data pages from the doublewrite

InnoDB: buffer...

InnoDB: Doing recovery: scanned up to log sequence number 9120034833

150125 16:12:51 InnoDB: Starting an apply batch of log records to the database...

InnoDB: Progress in percents: 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 150125 16:12:51 [ERROR] mysqld got signal 11 ;

This could be because you hit a bug. It is also possible that this binary

or one of the libraries it was linked against is corrupt, improperly built,

or misconfigured. This error can also be caused by malfunctioning hardware.

To report this bug, see http://kb.askmonty.org/en/reporting-bugs

We will try our best to scrape up some info that will hopefully help

diagnose the problem, but since we have already crashed, 

something is definitely wrong and this may fail.

Server version: 5.5.37-MariaDB-log

key_buffer_size=268435456

read_buffer_size=1048576

max_used_connections=0

max_threads=1002

thread_count=0

It is possible that mysqld could use up to 

key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 2332093 K bytes of memory

41 Hope that.

二分析

主要關注 mysqld got signal 11 的問題,從日志內(nèi)容分析來看,數(shù)據(jù)庫在機器crash 導致日志文件損壞,重啟之后無法正?；謴?更無法正常對外提供服務。

三解決

因為日志已經(jīng)損壞，這里采用非常規(guī)手段,首先修改innodb_force_recovery參數(shù)，使mysqld跳過恢復步驟，將mysqld 啟動,將數(shù)據(jù)導出來然后重建數(shù)據(jù)庫。

innodb_force_recovery可以設置為1-6,大的數(shù)字包含前面所有數(shù)字的影響。

1. (SRV_FORCE_IGNORE_CORRUPT):忽略檢查到的corrupt頁。
2. (SRV_FORCE_NO_BACKGROUND):阻止主線程的運行，如主線程需要執(zhí)行full purge操作，會導致crash。
3. (SRV_FORCE_NO_TRX_UNDO):不執(zhí)行事務回滾操作。
4. (SRV_FORCE_NO_IBUF_MERGE):不執(zhí)行插入緩沖的合并操作。
5. (SRV_FORCE_NO_UNDO_LOG_SCAN):不查看重做日志，InnoDB存儲引擎會將未提交的事務視為已提交。
6. (SRV_FORCE_NO_LOG_REDO):不執(zhí)行前滾的操作。

注意

a 當設置參數(shù)值大于0后，可以對表進行select,create,drop操作,但insert,update或者delete這類操作是不允許的。
b 當innodb_purge_threads 和 innodb_force_recovery一起設置會出現(xiàn)一種loop現(xiàn)象:

復制代碼代碼如下:

150125 17:07:42 InnoDB: Waiting for the background threads to start

150125 17:07:43 InnoDB: Waiting for the background threads to start

150125 17:07:44 InnoDB: Waiting for the background threads to start

150125 17:07:45 InnoDB: Waiting for the background threads to start

150125 17:07:46 InnoDB: Waiting for the background threads to start

150125 17:07:47 InnoDB: Waiting for the background threads to start

在my.cnf中修改以下兩個參數(shù)

復制代碼代碼如下:

innodb_force_recovery=6

innodb_purge_thread=0

重啟MySQL

復制代碼代碼如下:

150125 17:10:47 [Note] Crash recovery finished.

150125 17:10:47 [Note] Server socket created on IP: '0.0.0.0'.

150125 17:10:47 [Note] Event Scheduler: Loaded 0 events

150125 17:10:47 [Note] /vdata/webserver/mysql/bin/mysqld: ready for connections.

Version: '5.5.37-MariaDB-log' socket: '/tmp/mysql.sock' port: 3306 Source distribution

立即對數(shù)據(jù)庫做邏輯導出，完成之后將innodb_force_recovery設置為0 ，innodb_purge_thread=1 ,然后重建數(shù)據(jù)庫。
另外 MySQL 版本 5.5以及之前 ,當innodb_purge_threads =1，innodb_force_recovery >1 的情況會出現(xiàn)上文提到的循環(huán)報warning 問題（=1 沒有問題），

原因：

MySQL 的源代碼中顯示當innodb_purge_threads 和 innodb_force_recovery一起設置會出現(xiàn)loop循環(huán)

復制代碼代碼如下:

while (srv_shutdown_state == SRV_SHUTDOWN_NONE) { 

      if (srv_thread_has_reserved_slot(SRV_MASTER) == ULINT_UNDEFINED 

          || (srv_n_purge_threads == 1 

          && srv_thread_has_reserved_slot(SRV_WORKER) 

          == ULINT_UNDEFINED)) { 

          ut_print_timestamp(stderr); 

          fprintf(stderr, " InnoDB: Waiting for the background threads to start\n");

          os_thread_sleep(1000000); 

      } else { 

          break; 

      } 

  }

所以當需要設置innodb_force_recovery>1的時候需要關閉 innodb_purge_threads，設置為0（默認）。

四小結(jié)

MySQL crash 或者 MySQL 數(shù)據(jù)庫服務器 crash 會導致各種各樣的問題，比如主備之間的error 1594 (5.6 版本開啟crash-safe ，會最大程度上避免 error 1594的問題，以后會寫5.6新特性介紹該功能 )，error 1236，日志損壞，數(shù)據(jù)文件損壞，等等，本案例只是其中的一種，細心從日志中找的相關錯誤提示，逐步解決即可。

您可能感興趣的文章: