Oracle 11gR2启动ASM实例时遭遇ORA-29701

[grid@vm11gr2] /home/grid> sqlplus "/as sysasm"

    SQL*Plus: Release 11.2.0.1.0 Production on Sun Oct 25 10:16:21 2009

    Copyright (c) 1982, 2009, Oracle. All rights reserved.

    Connected to an idle instance.
    SQL> startup
    ORA-01078: failure in processing system parameters
    ORA-29701: unable to connect to Cluster Synchronization Service
    SQL>
    无法连接到CSS服务上.到操作系统上检查一下看看

    [grid@vm11gr2] /home/grid> crsctl check css
    CRS-4530: Communications failure contacting Cluster Synchronization Services daemon
    [grid@vm11gr2] /home/grid>
    [grid@vm11gr2] /home/grid> ps -ef|grep cssd
    果然没有CSS的服务daemon进程,再看一下HAS(High Availability Service)的状态

    [grid@vm11gr2] /home/grid> crsctl check has
    CRS-4638: Oracle High Availability Services is online

    [grid@vm11gr2] /home/grid> ps -ef|grep d.bin
    grid 5886 1 0 10:06 ? 00:00:01 /u01/app/grid/product/11.2/grid/bin/ohasd.bin reboot
    [grid@vm11gr2] /home/grid>
    发现HAS的服务确实启动了的,而ora.cssd和ora.diskmon这2个服务是依赖于HAS维护的.
   

    进一步查看各资源的状态
    [grid@vm11gr2] /home/grid> crs_stat -t
    Name                Type                 Target    State Host
    --------------------------------------------------------------
    ora.FLASH_DATA.dg   ora.diskgroup.type   OFFLINE OFFLINE vm11gr2
    ora.SYS_DATA.dg     ora.diskgroup.type   OFFLINE OFFLINE vm11gr2
    ora.asm             ora.asm.type         OFFLINE OFFLINE vm11gr2
    ora.cssd            ora.cssd.type        OFFLINE OFFLINE vm11gr2
    ora.diskmon         ora.diskmon.type     OFFLINE OFFLINE vm11gr2
    [grid@vm11gr2] /home/grid>
   
    [grid@vm11gr2] /home/grid> crsctl status resource -t
    --------------------------------------------------------------------------------
    NAME               TARGET          STATE        SERVER     STATE_      DETAILS
    --------------------------------------------------------------------------------
    Local Resources
    --------------------------------------------------------------------------------
    ora.FLASH_DATA.dg
                      OFFLINE          OFFLINE     vm11gr2
    ora.SYS_DATA.dg   

                      OFFLINE          OFFLINE     vm11gr2
    ora.asm           

                      OFFLINE          OFFLINE     vm11gr2
    --------------------------------------------------------------------------------
    Cluster Resources
    --------------------------------------------------------------------------------
    ora.cssd         

            1         OFFLINE          OFFLINE
    ora.diskmon      

            1         OFFLINE          OFFLINE

    再看一下ora.cssd和ora.diskmon的属性
    [grid@vm11gr2] /home/grid> crs_stat -p ora.cssd
    NAME=ora.cssd
    TYPE=ora.cssd.type
    ACTION_SCRIPT=
    ACTIVE_PLACEMENT=0
    AUTO_START=never
    CHECK_INTERVAL=30
    DESCRIPTION="Resource type for CSSD"
    FAILOVER_DELAY=0
    FAILURE_INTERVAL=3
    FAILURE_THRESHOLD=5
    HOSTING_MEMBERS=
    PLACEMENT=balanced
    RESTART_ATTEMPTS=5
    SCRIPT_TIMEOUT=600
    START_TIMEOUT=600
    STOP_TIMEOUT=900
    UPTIME_THRESHOLD=1m

    [grid@vm11gr2] /home/grid> crs_stat -p ora.diskmon
    NAME=ora.diskmon
    TYPE=ora.diskmon.type
    ACTION_SCRIPT=
    ACTIVE_PLACEMENT=0
    AUTO_START=never
    CHECK_INTERVAL=20
    DESCRIPTION="Resource type for Diskmon"
    FAILOVER_DELAY=0
    FAILURE_INTERVAL=3
    FAILURE_THRESHOLD=5
    HOSTING_MEMBERS=
    PLACEMENT=balanced
    RESTART_ATTEMPTS=10
    SCRIPT_TIMEOUT=60
    START_TIMEOUT=60
    STOP_TIMEOUT=60
    UPTIME_THRESHOLD=5s
    [grid@vm11gr2] /home/grid>
    到这里基本就找到了原因了,可以看到这两个资源的AUTO_START属性默认都设置为never,也就是说他们不会随着HAS服务的启动而自动启动的,尽管默认情况下HAS服务是开机自动启动的.好了，那我们就手动启动一下吧:
    [grid@vm11gr2] /home/grid> crsctl start resource ora.cssd
    CRS-2672: Attempting to start 'ora.cssd' on 'vm11gr2'
    CRS-2679: Attempting to clean 'ora.diskmon' on 'vm11gr2'
    CRS-2681: Clean of 'ora.diskmon' on 'vm11gr2' succeeded
    CRS-2672: Attempting to start 'ora.diskmon' on 'vm11gr2'
    CRS-2676: Start of 'ora.diskmon' on 'vm11gr2' succeeded
    CRS-2676: Start of 'ora.cssd' on 'vm11gr2' succeeded
    [grid@vm11gr2] /home/grid>

    注:ora.cssd和ora.diskmon这两个服务是有依赖关系的，启动哪个都会把两个都起来.

    [grid@vm11gr2] /home/grid> crs_stat -t
    Name                Type                 Target    State Host
    --------------------------------------------------------------
    ora.FLASH_DATA.dg   ora.diskgroup.type   OFFLINE OFFLINE vm11gr2
    ora.SYS_DATA.dg     ora.diskgroup.type   OFFLINE OFFLINE vm11gr2
    ora.asm             ora.asm.type         OFFLINE OFFLINE vm11gr2
    ora.cssd            ora.cssd.type        ONLINE  ONLINE  vm11gr2
    ora.diskmon         ora.diskmon.type     ONLINE  ONLINE  vm11gr2
    [grid@vm11gr2] /home/grid>
    CSS服务起来了，重启动asm instance   

    [grid@vm11gr2] /home/grid> sqlplus "/as sysasm"

    SQL*Plus: Release 11.2.0.1.0 Production on Sun Oct 25 10:30:03 2009

    Copyright (c) 1982, 2009, Oracle. All rights reserved.

    Connected to an idle instance.

    SQL> startup
    ASM instance started

    Total System Global Area 284565504 bytes
    Fixed Size 1336036 bytes
    Variable Size 258063644 bytes
    ASM Cache 25165824 bytes
    ASM diskgroups mounted
    SQL> exit
    Disconnected from Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - Production
    With the Automatic Storage Management option
    [grid@vm11gr2] /home/grid> crs_stat -t
    Name                Type                 Target    State Host
    --------------------------------------------------------------
    ora.FLASH_DATA.dg   ora.diskgroup.type   ONLINE  ONLINE  vm11gr2
    ora.SYS_DATA.dg     ora.diskgroup.type   ONLINE  ONLINE  vm11gr2
    ora.asm             ora.asm.type         ONLINE  ONLINE  vm11gr2
    ora.cssd            ora.cssd.type        ONLINE  ONLINE  vm11gr2
    ora.diskmon         ora.diskmon.type     ONLINE  ONLINE  vm11gr2
    [grid@vm11gr2] /home/grid>

    tips

      1)默认情况下HAS(High Availability Service)是自动启动的.通过如下命令可以取消和启用自动启动
        crsctl disable has
        crsctl enable has
      2)HAS手动启动和停止
        crsctl start has
        crsctl stop has
      3)查看HAS的状态
        crsctl check has
      4)如果想让ora.css和ora.diskmon服务随着HAS的启动而自动启动，那么你可以这两个服务的AUTO_START属性
        crsctl modify resource "ora.cssd" -attr "AUTO_START=1"
        or
        crsctl modify resource "ora.diskmon" -attr "AUTO_START=1"
      5)如果想取消ora.css和ora.diskmon的Auto start
        crsctl modify resource "ora.cssd" -attr "AUTO_START=never"
        crsctl modify resource "ora.diskmon" -attr "AUTO_START=never"