11gR2启动ASM实例时遭遇ORA-29701

一、引言

    前几天安装了Oracle 11gR的ASM，但是今天想要启动ASM的时候遇到下面的情况：

1 [grid@yft ~]$ sqlplus /nolog
2
3 SQL*Plus: Release 11.2.0.1.0 Production on Sun Mar 24 08:37:57 2013
4
5 Copyright (c) 1982, 2009, Oracle.  All rights reserved.
6
7 SQL> conn / as sysasm
8 Connected to an idle instance.
9 SQL> startup
10 ORA-01078: failure in processing system parameters
11 ORA-29701: unable to connect to Cluster Synchronization Service

二、解决问题

    无法连接到CSS服务上，到操作系统上检查一下：

1 [grid@yft ~]$ crsctl check css
2 CRS-4530: Communications failure contacting Cluster Synchronization Services daemon
3 [grid@yft ~]$ ps -ef |grep cssd
4 grid      5327  4963  0 08:33 pts/0    00:00:00 grep cssd

    果然没有CSS的服务daemon进程，再看一下HAS(High Availability Service)的状态：

1 [grid@yft ~]$ crsctl check has
2 CRS-4638: Oracle High Availability Services is online
3 [grid@yft ~]$ ps -ef |grep d.bin
4 grid      4704     1  0 08:22 ?        00:00:02 /u01/app/grid/product/11.2.0/grid/bin/ohasd.bin reboot
5 grid      5334  4963  0 08:34 pts/0    00:00:00 grep d.bin

    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：

1 [grid@yft ~]$ crsctl status resource -t
2 --------------------------------------------------------------------------------
3 NAME           TARGET  STATE        SERVER                   STATE_DETAILS
4 --------------------------------------------------------------------------------
5 Local Resources
6 --------------------------------------------------------------------------------
7 ora.DATA.dg
8                OFFLINE OFFLINE      yft
9 ora.DATA1.dg
10                OFFLINE OFFLINE      yft
11 ora.asm
12                OFFLINE OFFLINE      yft
13 --------------------------------------------------------------------------------
14 Cluster Resources
15 --------------------------------------------------------------------------------
16 ora.cssd
17       1        ONLINE  OFFLINE
18 ora.diskmon
19       1        ONLINE  OFFLINE
20 ora.orcl.db
21       1        OFFLINE OFFLINE

    再看一下ora.cssd和ora.diskmon的属性：

1 [grid@yft ~]$ crs_stat -p ora.cssd
2 NAME=ora.cssd
3 TYPE=ora.cssd.type
4 ACTION_SCRIPT=
5 ACTIVE_PLACEMENT=0
6 AUTO_START=never
7 CHECK_INTERVAL=30
8 DESCRIPTION="Resource type for CSSD"
9 FAILOVER_DELAY=0
10 FAILURE_INTERVAL=3
11 FAILURE_THRESHOLD=5
12 HOSTING_MEMBERS=
13 PLACEMENT=balanced
14 RESTART_ATTEMPTS=5
15 SCRIPT_TIMEOUT=600
16 START_TIMEOUT=600
17 STOP_TIMEOUT=900
18 UPTIME_THRESHOLD=1m
19
20 [grid@yft ~]$ crs_stat -p ora.diskmon
21 NAME=ora.diskmon
22 TYPE=ora.diskmon.type
23 ACTION_SCRIPT=
24 ACTIVE_PLACEMENT=0
25 AUTO_START=never
26 CHECK_INTERVAL=20
27 DESCRIPTION="Resource type for Diskmon"
28 FAILOVER_DELAY=0
29 FAILURE_INTERVAL=3
30 FAILURE_THRESHOLD=5
31 HOSTING_MEMBERS=
32 PLACEMENT=balanced
33 RESTART_ATTEMPTS=10
34 SCRIPT_TIMEOUT=60
35 START_TIMEOUT=60
36 STOP_TIMEOUT=60
37 UPTIME_THRESHOLD=5s

    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，现在手动启动一下：

1 [grid@yft ~]$ crsctl start resource ora.cssd
2 CRS-2672: Attempting to start 'ora.cssd' on 'yft'
3 CRS-2679: Attempting to clean 'ora.diskmon' on 'yft'
4 CRS-2681: Clean of 'ora.diskmon' on 'yft' succeeded
5 CRS-2672: Attempting to start 'ora.diskmon' on 'yft'
6 CRS-2676: Start of 'ora.diskmon' on 'yft' succeeded
7 CRS-2676: Start of 'ora.cssd' on 'yft' succeeded
8
9 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。
10
11 [grid@yft ~]$ crs_stat -t
12 Name           Type           Target    State     Host
13 ------------------------------------------------------------
14 ora.DATA.dg    ora....up.type OFFLINE   OFFLINE
15 ora.DATA1.dg   ora....up.type OFFLINE   OFFLINE
16 ora.asm        ora.asm.type   OFFLINE   OFFLINE
17 ora.cssd       ora.cssd.type  ONLINE    ONLINE    yft
18 ora.diskmon    ora....on.type ONLINE    ONLINE    yft
19 ora.orcl.db    ora....se.type OFFLINE   OFFLINE

    CSS服务起来了，重启动asm instnace

1 [grid@yft ~]$ crs_stat -t
2 Name           Type           Target    State     Host
3 ------------------------------------------------------------
4 ora.DATA.dg    ora....up.type ONLINE    ONLINE    yft
5 ora.DATA1.dg   ora....up.type ONLINE    ONLINE    yft
6 ora.asm        ora.asm.type   ONLINE    ONLINE    yft
7 ora.cssd       ora.cssd.type  ONLINE    ONLINE    yft
8 ora.diskmon    ora....on.type ONLINE    ONLINE    yft
9 ora.orcl.db    ora....se.type OFFLINE   OFFLINE

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动自动启动

    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.cssd和ora.diskmon服务随着HAS的启动而自动启动，那么你可以把这两个服务的AUTO_START属性：

    crsctl modify resource "ora.cssd" -attr "AUTO_START=1"

    crsctl modify resource "ora.diskmon" -attr "AUTO_START=1"

5)、如果想取消ora.cssd和ora.diskmon的Auto start

    crsctl modify resource "ora.cssd" -attr "AUTO_START=never"

    crsctl modify resource "ora.diskmon" -attr "AUTO_START=never"