ITPub博客

今天上午，公司的一台数据库HDIC的其中一个磁盘空间满了，导致归档失败，外部程序无法连入。经过详细检查，发现是由于设置了强制的恢复点（v$restore_point）造成的，于是删除掉不用的恢复点。但是现在还是出现了磁盘空间大量占用的情况，经检查，是DATABASE FLASHBACK LOG占用的，用了1.4TB，挺吓人的。查询v$flashback_database_logfile确定文件名后，进入ASMCMD手工删了一些。但是v$flashback_database_logfile中还有记录，挺不爽的，做事就要做得干净彻底嘛。经过思考和实践，发现了删除v$flashback_database_logfile记录的方法，其实

[oracle@bj-etlsrv1 dump_etl]$ oerr ora 3921339213, 00000, "Metadata processing is not available"// *Cause: The Data Pump could not use the Metadata API. Typically,// this is caused by the XSL stylesheets not being set up properly.// *Action: Connect AS SYSDBA and execute dbms_metadata_util.load_stylesheets// to reload the stylesheets.------解决办法：用SYS用户执行dbms_metadata_util.load_stylesheets即可解决.[@more@]

refs:http://blog.csdn.net/daimin1983/article/details/4291661http://space.itpub.net/index.php?action/viewspace/itemid/629063---------------TNS-12547 Linux Error: 104: Connection reset by peer解决过程参考:http://blog.chinaunix.net/u/7121/showart_403812.html[oracle@MyMachine log]$ lsnrctl startLSNRCTL for Linux: Version 10.2.0.1.0 - Production on 23-JUN-2009 09:53:26Copyright (c) 1991, 2005, Oracle. All rights reserved.Starting /opt/oracle/product/10.2.1/db_1/bin/tnslsnr: please wait...TNSLSNR for

库启动正常，以前用作迁移开发。启动后，服务器本地可以连上，但是远程不行，报TNS-12523 。后台日志显示出 以前配置的网关连接失效。后来实在没招了，设置了本地监听参数，再试，竟然好了。唉，真神奇。。。alter system set local_listener='(DESCRIPTION=(ADDRESS_LIST=(ADDRESS=(PROTOCOL=TCP)(HOST=172.16.5.61)(PORT=1521))))' scope=both ;[@more@]

1G里有个wait event 叫asynch descriptor resize ORACLE metalink里并没有描述 根据国外大神的研究 发现这个参数是 ORACLE 在调用direct I/O的时候。会调用OS 的syscall 来获取 AIO 的descriptors 。4096 descriptors per process 当你的执行结束的时候 ORACLE就会对其进行释放 于是就产生了 拿来 放下 拿来放下 的动作。 如果发现了这样的问题 那么首先请检查 是否超过1%的占用时间 在你的系统里 。如果不超过 那么无视他吧。如果超过了20% 那么你或许要考虑 SR了。 如果在1-20之间 首先请先考虑你的I/O子系统的问题 至此有个疑问 如

-- 错误代码举例:(from alert_DBSID.log)Tue Mar 15 18:35:21 2011+188791 Thread 1 advanced to log sequence 783 (LGWR switch)+188792 Current log# 3 seq# 783 mem# 0: /opt/app/oracle/oradata/Homelink/redo03.log+188793 Thread 1 cannot allocate new log, sequence 784+188794 Checkpoint not complete+188795 Current log# 3 seq# 783 mem# 0: /opt/app/oracle/oradata/Homelink/redo03.log+188796 Thread 1 advanced to log sequence 784 (LGWR switch)+188797 Current log# 1 seq# 784 mem# 0: /opt/app/oracle/oradata

--错误代码举例: (from alert_DBSID.log)Tue Mar 22 17:20:29 2011+210778 Process startup failed, error stack:+210779 Errors in file /opt/app/oracle/diag/rdbms/homelink/Homelink/trace/Homelink_psp0_3080224.trc:+210780 ORA-27300: OS system dependent operation:fork failed with status: 12+210781 ORA-27301: OS failure message: Not enough space+210782 ORA-27302: failure occurred at: skgpspawn3-- 解决: 增加虚拟内存.减少SGA,PGA[@more@]

Link:http://sadhus.blog.sohu.com/91144960.htmlsummary:这是和网络连接相关的一个错误,Metalink上给出了如下的解决方案:1.set INBOUND_CONNECT_TIMEOUT_listener_name =0 in listener.ora2. set SQLNET.INBOUND_CONNECT_TIMEOUT = 0 in sqlnet.ora of server.3. stop and start both listener and database.4. Now try to connect to DB and observe the behaviour[@more@]

使用SKIP_CORRUPT_BLOCKS或Event 10231来提取数据http://blog.chinaunix.net/u/12960/showart_229495.html[@more@]

C:Documents and SettingsAdministrator>lsnrctl startLSNRCTL for 32-bit Windows: Version 11.2.0.1.0 - Production on 23-DEC-2010 11:28:06Copyright (c) 1991, 2010, Oracle. All rights reserved.Starting tnslsnr: please wait...TNSLSNR for 32-bit Windows: Version 11.2.0.1.0 - ProductionSystem parameter file is E:appAdministratorproduct11.2.0dbhome_1networkadminlistener.oraLog messages written to e:appadministratorproduct11.2.0dbhome_1logdiagtnslsnrPC-20091017114alisteneralertlog.xmlError listening o

10g中的transaction等待事件http://empo007.itpub.net/post/12076/496412[@more@]

************error stack ****************C:Documents and SettingsAdministrator>oradim -startup -sid emrDIM-04503: Message 4503 not found; product=RDBMS; facility=ORADIM*******************************this error is similar to sqlplus which refer to mssage file. resolution: 1, set ORACLE_HOME to the correct directory .2, set the environment PATH variable to the correct %oracle_home%bin value. [@more@]

*********error message stack ************C:Documents and SettingsAdministrator>sqlplus / as sysdbaError 14 initializing SQL*PlusMessage file sp1<lang>.msb not foundSP2-0750: You may need to set ORACLE_HOME to your Oracle software directory******************************resolution: 1, set ORACLE_HOME to the correct directory .2, set the environment PATH variable to the correct %oracle_home%bin value. [@more@]

10g用户故障快速恢复方法研究http://www.docin.com/p-22877284.htmloracle10g Flashback Databasehttp://singlelove1983.blog.163.com/blog/static/50849047200801794342791/使用Oracle10g Flashback database功能恢复用户错误http://www.eygle.com/archives/2005/03/eoaoracle10g_fl.html[@more@]

direct path read/write等待的分析http://iamwang.blog.sohu.com/93404343.html[@more@]

处理数据库慢的一般方法 以下在XZH2000 博客摘取如果不用分析工具如statspack等，如何快速找出原因？lcm_0090828给出自己常用的checklist:1. 检查表空间资源2. 归档是否正常（归档空间满时，会等待归档，之后用户进程就死掉了）3. 检查alert 信息4. 用vmstat, top命令看看系统i/o, cpu的繁忙程度，以及哪个进程在消耗资源，结合v$process, v$session, v$sqlarea 等视图查找出终端机器，以及其使用的sql 语句等。5. 有没有死锁6. 有时用户进程运行的程序会报错，就根据错误提示寻找根源及解决办法，可能是初始化参数设置不合理[@more@]

-- 使用了DB VAULT之后，导出数据出错，SYS用户竟然也权限不足。E:>expdp '/ as sysdba ' tables=scott.test directory=dmpdir dumpfile=test04.dmp COMPRESSION=ALLExport: Release 11.2.0.1.0 - Production on 星期四 11月 25 18:31:31 2010Copyright (c) 1982, 2009, Oracle and/or its affiliates. All rights reserved.连接到: Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - ProductionWith the Partitioning, Oracle Label Security, OLAP, Data Mining,Oracle Database Vault and Real Application Testing op