ORA-3113 when attempting to STARTUP Oracle

3113 我也遇见过，数据库启动不同阶段，有不同解决办法．metalink写的很详细了

(A) ORA-3113 when attempting to STARTUP Oracle
   
  There are several phases involved in starting up a database. If ORA-3113
  occurs during startup then abort the instance and start up using the
  sequence below. If an error occurs at any step then see the related notes
  below.

        a. Start any required services.                On error see A1
           Eg: On NT start the OracleServiceSID

        b. Connect as a SYSDBA user.                  On error see A1
           Eg:  sqlplus /nolog
                SQL> connect / as sysdba

        c. Startup nomount.                          On error see A1
           Eg:
                SQL> startup nomount

        d. Mount the database.                         On error see A1 and A2
           Eg:
                SQL> alter database mount;

        e. Recover the database                 On error see A3
           Eg:
                SQL> recover database

        f. Open the database                         On error see A4
           Eg:        
                SQL> alter database open;

        g. Wait 3 minutes then issue a select.   On error see A4
           Eg:        
                SQL> select count(*) from DBA_OBJECTS;



A1) Errors connecting as SYSDBA / Internal OR on startup nomount
       
        There is something fundamental wrong with the software / environment
        if you cannot connect to Server Manager as a DBA user.
        The steps here cover errors such as ORA-3113, ORA-12547: TNS:lost contact
        or similar errors connecting to Oracle or starting the instance NOMOUNT.  
        Check the following items:

        A1.1)        If possible reboot the server disabling any automatic
                startup of Oracle before you do so. This may seem drastic
                but helps make sure you are working from a consistent
                starting point.

        A1.2)         Check your environment points at the expected ORACLE_HOME
                and ORACLE_SID and that TWO_TASK is not set (Unix) or
                LOCAL is not set (NT registry).
                Check the USER_DUMP_DEST and BACKGROUND_DUMP_DEST and default
                trace directories under this environment for any user trace
                files or alert log entries generated. These may help indicate
                the cause of the problem.
                Eg: ORA-600[SKGMINVALID] may indicate a problem with the
                    shared memory Unix parameters on Unix systems.
                Try to show that any trace file / alert log entry you
                find is truely related to the "CONNECT" command by re-issuing
                the "connect" and checking for a new trace file / alert entry
                at the time of the error.

        A1.3)   Unix only:
                Some Unix platforms need LD_LIBRARY_PATH to be set
                correctly to resolve any dynamically linked libraries.
                As the user with the problem:

                        % script. /tmp/ldd.out
                        % id
                        % cd $ORACLE_HOME/bin
                        % ldd oracle
                        % exit

                If the 'ldd' command does not exist go to the next step below.
                Check that all lines listed show a full library file. If there
                are any 'not found' lines reported contact Oracle support
                with the output of /tmp/ldd.out .


        A1.4)   Unix only:
                Your 'oracle' executable may be corrupt. Relink it thus:

                        Log in as the 'oracle' user.
                        % script. /tmp/relink.out
                        % cd $ORACLE_HOME/rdbms/lib
                        % mv $ORACLE_HOME/bin/oracle $ORACLE_HOME/bin/oracle.dd.mon.yy
                        % rm -f ./oracle        
                        % make -f ins_rdbms.mk ioracle
                        % exit

                Prior to Oracle7.3 the relink command was:
                        make -f oracle.mk ioracle

                If this reports any errors Oracle support will need to see
                the contents of the file /tmp/relink.out .

       
        A1.5)   Have you installed the Parallel Server Option ?
                ORA-3113 can occur if you have installed the Parallel
                Server Option but do NOT have a Distributed Lock Manager
                installed or running correctly.

                Unix:
                  If the Parallel Server Option was installed by accident
                  then it can be de-installed by relinking.
                  Eg:
                        Shut down any Oracle instances
                        % script. /tmp/relink.out
                        % cd $ORACLE_HOME/rdbms/lib
                        # 'oracle' should not exist so delete it if it present
                        % rm -f oracle
                        % make -f ins_rdbms.mk no_parropt ioracle
                        % exit

                  NB: Do NOT deinstall the Parallel Server Option if the
                      database is using Parallel Server unless both nodes
                      are shut down otherwise database corruption could occur.

        A1.6)   If the error is on STARTUP NOMOUNT:

                        Check the init.ora file used to start the database.
                        This provides the configuration details used
                        configure the instance. To help isolate the problem
                        it may be useful use a very basic init.ora file
                        when starting the instance. If this works then
                        parameters can be increased / introduced one at a
                        time to see if there is a problem with a particular
                        setting.

        A1.7)         Check for server side trace files which may give more
                indication what the underlying problem is.
                See section C for details on how to check
                for server trace files.
               

        A1.8)         Ensure there is free disk space in:

                  a. Your USER_DUMP_DEST and BACKGROUND_DUMP_DEST locations

                  b. Your AUDIT destination (Unix)
                        The default is $ORACLE_HOME/rdbms/audit

                  c. Your Oracle Trace directory if Oracle Trace is enabled
                        See Note 45482.1

        A1.9) Windows 2000 - Only
              If the Server's sqlnet.ora file contains Authnetication services
              which are NOT reachable by Oracle, then an ORA-3113 error will
              result.
              For example, if the sqlnet.ora file contains the parameter:
              SQLNET.AUTHENTICATION_SERVICES = (NTS) and the Oracle database
              is moved from a Windows NT Domain to an Active Directory one,
              or if a Domain Controller is introduced, then an error will
              result trying to start the database.
              Remove the sqlnet.authentication_services line so that Oracle
              does not look for a non-existent KDC (Kerberos Domain Controller).


                       
A2) Errors Mounting the database

        Check all the items in A1 first.

        If an error occurs when mounting the database there may be problems
        with the control-files or data files, or with resources required to
        open these files.

        A2.1)         The location of the control files are specified in the
                init.ora file.  Try mounting using each controlfile in
                turn.
                eg: "Shutdown abort",
                    edit the init.ora to refer to ONE of the controlfiles only,
                    "startup nomount",
                    "alter database mount"
                Repeat for each controlfile to see if any controlfile works.
               
        A2.2)        It is possible to re-create the controlfiles if you know the
                location of all datafiles and online logs, or to restore an old
                backup controlfile. Always back up the current controlfiles before
                restoring any backup copies or issuing a CREATE CONTROLFILE
                command.
                The steps for this are not documented here.

        A2.3)   Unix:
                Some unix platforms have a 'truss' command (or 'tusc').
                If available this can be used to help trace how far Oracle
                gets before the error occurs.
                Eg:  
                        % truss -o /tmp/truss.out -f svrmgrl

                Keep the file /tmp/truss.out safe - Oracle Support MAY need to see it.


A3) Errors on RECOVER DATABASE

          ORA-3113 during recover database is often related to a corruption on the
        database or redo stream which causes the shadow process to die. There should
        be a server side trace file produced for this sort of problem.
        See Section C for details on how to locate any trace files
        from both USER_DUMP_DEST and BACKGROUND_DUMP_DEST.

        A3.1)        If the "recover database" fails fairly quickly then it
                may help to collect the redo up to the point of failure as this
                may help identify where the problem is.

                Use the following commands just prior to the RECOVER DATABASE
                command:

                  SQL> alter session set max_dump_file_size=unlimited;
                  SQL> alter session set events
                  2> '10228 trace name context forever, level 10';
                  SQL> RECOVER DATABASE
               
                This causes redo information to be written to the user trace
                file. The last items of redo may help determine which file
                has problems.

        A3.2)        If you do not have many datafiles in the database it may be
                as quick to recover each file in turn to see if this narrows
                down the problem.
                Eg:
                  SQL> select name from v$datafile;
               
                and then for each file:

                  SQL> RECOVER DATAFILE 'full_file_name'

                If this gets to a problem file then back up the file and
                use standard recovery options as if the file was lost.
               

A4) Errors on ALTER DATABASE OPEN

        Database open performs very many operations and so it is necessary
        to collect any trace information before determining the next steps.
        However, the following may help isolate the problem more quickly:

        A4.1)         Move files out of your USER_DUMP_DEST and BACKGROUND_DUMP_DEST
                directory as these steps will generate a lot of trace.

        A4.2)         Edit the init.ora file and add the lines:

                        event="10046 trace name context forever, level 12"
                        event="10015 trace name context forever, level 1"
                        event="10228 trace name context forever, level 1"

                If you already have "EVENT=" lines in the init.ora file
                this MUST directly below the other "Event=" lines.
                These lines will trace:

                        SQL and BIND activity during startup
                        REDO applied
                        Information about transactional rollback required

        A4.3)         Startup the instance as described at the top of this section.
               
                As soon as the error occurs REMOVE the above events from the init.ora
                file and shutdown. Collect together the trace files and alert logs
                as described in Section C

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/8102208/viewspace-483886/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/8102208/viewspace-483886/