sub send_mail { my($sendername,$receivename,$mailfrom,$mailto,$subject,$content) = @_; // $sendername,$receivename,$subject,$content都有可能是中文。 my $mailhost=‘mail host'; my $smtp = Net::SMTP->new( Host    =>  $mailhost, Hello   =>  $mailhost, Timeout =>  30, Debug   =>  1 ); $smtp-

USERLOCATION 表结构。 mysql>show create table USERLOCATION\G *************************** 1. row *************************** Table: USERLOCATION Create Table: CREATE TABLE `USERLOCATION` ( `ID_` bigint(20) NOT NULL AUTO_INCREMENT, `LOCATIONID_` bigint(20) DEFAULT '0', `LOCATIONTYPE_` smallint(6) DEFAULT NULL, `RELATIONSHIPTYPE_` smallint(6) DEFAULT NULL, `USERID_` bigint(20) DEFAULT '0', `CREATED_` datetime DEFAU

091002 18:13:57 mysqld_safe Starting mysqld daemon with databases from /massdata/dashboard/data 091002 18:14:00  InnoDB: Started; log sequence number 251 4124007832 InnoDB: !!! innodb_force_recovery is set to 4 !!! 091002 18:14:00 [Note] Event Scheduler: Loaded 0 events 091002 18:14:00 [Note] /usr/sbin/mysqld: ready for connections. Version: '5.1.33'  socket: '/massdata/dashboard/data/mysql.sock'  port: 3306  SUSE MySQL RPM 091002 18:14:15  Inno

*** glibc detected *** top: realloc(): invalid next siz

1.mysql 从库上出的错，这个是把master的binglog 设置成mixed模式，触发器造成的slave中断。 | Waiting for master to send event | 192.168.100.149 | repl | 3306 | 60 | mysql-bin.000132 | 206432547 | wfn-db-1-relay-bin.000023 | 182939066 | mysql-bin.000131 | Yes | No | m4n | information_schema,mysql,wtst,wfnphoto | | | | |

Linux文件预读算法磁盘I/O性能的发展远远滞后于CPU和内存，因而成为现代计算机系统的一个主要瓶颈。预读可以有效的减少磁盘的寻道次数和应用程 序的I/O等待时间，是改进磁盘读I/O性能的重要优化手段之一。本文作者是中国科学技术大学自动化系的博士生，他在1998年开始学习Linux，为了 优化服务器的性能，他开始尝试改进Linux kernel，并最终重写了内核的文件预读部分，这些改进被收录到Linux Kernel 2.6.23及其后续版本中。 从寄存器、L1/L2高速缓存、内存、闪存，到磁盘/光盘/磁带/存储网络，计算机的各级存储器硬件组 成了一

One of very interesting techniques Innodb uses is technique called “doublewrite” It means Innodb will write data twice when it performs table space writes – writes to log files are done only once. So why doublewrite is needed ? It is needed to archive data safety in case of partial page writes. Innodb does not log full pages to the log files, but uses what is called “physiological” logging which means log records c

echo "i am hifdafafdst"|awk '{gsub(/am/,"abcc",$0);print $0}' i abcc hifdafafdst

InnoDB引擎对于巨量数据有强大的优势。特别是ROW LOCK的设计，即使多线程同时修改表中的多处数据，也不会造成锁死和等待。好处我就不多说了。 但是如果在InnoDB中大量使用 INSERT…SELECT这种批量插入的语句，InnoDB的优势便很遗憾的无法发挥了。为什么？因为MySQL为了保证Log和 Replication阵列的正常工作，就必须按顺序记录下所有数据插入记录，以保证当数据库阵列中的其他数据库通过Log建立数据库镜像的正确性。其前 提就是记录中的数据必须是确定的数值，而不能是不确定的数据。INSERT…SE

在show innodb status　的结果中有这样一段信息。 有些mysql系统中hash searches/s　这个值很高，有些mysql系统中就不会很高。 INSERT BUFFER AND ADAPTIVE HASH INDEX ------------------------------------- Ibuf: size 1, free list len 0, seg size 2, 0 inserts, 0 merged recs, 0 merges Hash table size 34679, used cells 0, node heap has 1 buffer(s) 0.00 hash searches/s, 0.00 non-hash searches/s 官方手册中有这样的

InnoDB: End of page dump 090626 9:14:39 InnoDB: Page checksum 1782824756, prior-to-4.0.14-form checksum 117720807 InnoDB: stored checksum 517774722, prior-to-4.0.14-form stored checksum 117720807 InnoDB: Page lsn 22 3649314269, low 4 bytes of lsn at page end 3649314269 InnoDB: Page number (if stored to page already) 211, InnoDB: space id (if created with >= MySQL-4.1.1 and stored already) 420 InnoDB: Page may be an update undo log page InnoDB: Page may be an in

加为好友

已有人次访问本空间   什么是RSS？

wanghusr	Owen26813
avauntage	szrednick
ismurray	thinkinginlamp
23460724	ysihai88