在Grid
Infrastructure软件安装的完成后,系统会提示在所有的rac节点上面运行脚本root.sh,通过查看root.sh的运行日志就可以分析出oracle
GI群集的整个启动过程。默认情况下这个文件位于$GI_HOME的cfgtoollogs下面,文件名称格式是rootcrs_.log。
这个脚本运行先通过读取安装过程中输入的各种配置信息(比如ASM DISK/ASM DISK
GROUP/VIP/SCAN/CLUSTER NAME),然后注册必要的oracle
cluster_resource的type,
crsctl add type ora.daemon.type -basetype cluster_resource
crsctl add type ora.haip.type -basetype cluster_resource
crsctl add type ora.mdns.type -basetype ora.daemon.type
crsctl add type ora.gpnp.type -basetype ora.daemon.type
crsctl add type ora.gipc.type -basetype ora.daemon.type
crsctl add type ora.cssd.type -basetype ora.daemon.type
crsctl add type ora.cssdmonitor.type -basetype
ora.daemon.type
crsctl add type ora.crs.type -basetype ora.daemon.type
crsctl add type ora.evm.type -basetype ora.daemon.type
crsctl add type ora.ctss.type -basetype ora.daemon.type
crsctl add type ora.crf.type -basetype ora.daemon.type
crsctl add type ora.asm.type -basetype ora.daemon.type
crsctl add type ora.drivers.acfs.type -basetype
ora.daemon.type
crsctl add type ora.diskmon.type -basetype ora.daemon.type
然后把添加这些cluster_resource到群集中
crsctl add resource ora.mdnsd
crsctl add resource ora.gpnpd
crsctl add resource ora.gipcd
crsctl add resource ora.diskmon
crsctl add resource ora.cssdmonitor
crsctl add resource ora.cssd
crsctl add resource ora.ctssd
crsctl add resource ora.cluster_interconnect.haip
crsctl add resource ora.evmd
crsctl add resource ora.asm
crsctl add resource ora.crsd
接下来运行命令:
crsctl start resource ora.cssd -init -env CSSD_MODE=-X
这个命令就是要以eXclusive模式启动cssd服务,在这个服务启动后,创建ASM磁盘组,然后创建voting
disk
这个工作完成后,在关闭群集服务,然后再逐个打开,下面是打开的详细日志:
2012-09-01 18:19:53: Executing cmd:
/u01/app/11.2.0/grid/bin/crsctl start
resource ora.mdnsd -init
2012-09-01 18:19:55: Command output:
> CRS-2672: Attempting to start
'ora.mdnsd' on 'gridnode1'
> CRS-2676: Start of 'ora.mdnsd'
on 'gridnode1' succeeded
>End Command output
2012-09-01 18:19:55: Start of resource "ora.mdnsd" Succeeded
2012-09-01 18:19:55: Executing cmd:
/u01/app/11.2.0/grid/bin/crsctl start
resource ora.gpnpd -init
2012-09-01 18:19:56: Command output:
> CRS-2672: Attempting to start
'ora.gpnpd' on 'gridnode1'
> CRS-2676: Start of 'ora.gpnpd'
on 'gridnode1' succeeded
>End Command output
2012-09-01 18:19:56: Start of resource "ora.gpnpd" Succeeded
2012-09-01 18:19:56: Executing cmd:
/u01/app/11.2.0/grid/bin/gpnptool lfind
2012-09-01 18:19:57: Command output:
> Success. Local gpnpd
found.
>End Command output
2012-09-01 18:19:57: Successfully started requested Oracle stack
daemons
2012-09-01 18:19:57: Starting CSS in clustered mode
2012-09-01 18:19:57: Executing cmd:
/u01/app/11.2.0/grid/bin/crsctl start
resource ora.cssd -init
2012-09-01 18:22:27: Command output:
> CRS-2672: Attempting to start
'ora.cssdmonitor' on 'gridnode1'
> CRS-2672: Attempting to start
'ora.gipcd' on 'gridnode1'
> CRS-2676: Start of
'ora.cssdmonitor' on 'gridnode1' succeeded
> CRS-2676: Start of 'ora.gipcd'
on 'gridnode1' succeeded
> CRS-2672: Attempting to start
'ora.cssd' on 'gridnode1'
> CRS-2672: Attempting to start
'ora.diskmon' on 'gridnode1'
> CRS-2676: Start of 'ora.diskmon'
on 'gridnode1' succeeded
> CRS-2676: Start of 'ora.cssd' on
'gridnode1' succeeded
>End Command output
2012-09-01 18:22:27: Executing cmd: /u01/app/11.2.0/grid/bin/crsctl
check css
2012-09-01 18:22:28: Command output:
> CRS-4529: Cluster
Synchronization Services is online
>End Command output
2012-09-01 18:22:28: CRS-2672: Attempting to start
'ora.cssdmonitor' on 'gridnode1'
2012-09-01 18:22:28: CRS-2672: Attempting to start 'ora.gipcd' on
'gridnode1'
2012-09-01 18:22:28: CRS-2676: Start of 'ora.cssdmonitor' on
'gridnode1' succeeded
2012-09-01 18:22:28: CRS-2676: Start of 'ora.gipcd' on 'gridnode1'
succeeded
2012-09-01 18:22:28: CRS-2672: Attempting to start 'ora.cssd' on
'gridnode1'
2012-09-01 18:22:28: CRS-2672: Attempting to start 'ora.diskmon' on
'gridnode1'
2012-09-01 18:22:28: CRS-2676: Start of 'ora.diskmon' on
'gridnode1' succeeded
2012-09-01 18:22:28: CRS-2676: Start of 'ora.cssd' on 'gridnode1'
succeeded
2012-09-01 18:22:28: Executing cmd:
/u01/app/11.2.0/grid/bin/crsctl start
resource ora.ctssd -init -env
USR_ORA_ENV=CTSS_REBOOT=TRUE
2012-09-01 18:22:30: Command output:
> CRS-2672: Attempting to start
'ora.ctssd' on 'gridnode1'
> CRS-2676: Start of 'ora.ctssd'
on 'gridnode1' succeeded
>End Command output
2012-09-01 18:22:30: Start of resource "ora.ctssd" Succeeded
2012-09-01 18:22:30: CHM/OS is not supported on this platform
2012-09-01 18:22:30: start_clster: Failed to start CRF
2012-09-01 18:22:30: Executing cmd:
/u01/app/11.2.0/grid/bin/crsctl start
resource ora.cluster_interconnect.haip -init
2012-09-01 18:22:39: Command output:
> CRS-2672: Attempting to start
'ora.cluster_interconnect.haip' on 'gridnode1'
> CRS-2676: Start of
'ora.cluster_interconnect.haip' on 'gridnode1' succeeded
>End Command output
2012-09-01 18:22:39: Start of resource
"ora.cluster_interconnect.haip" Succeeded
2012-09-01 18:22:39: Executing cmd:
/u01/app/11.2.0/grid/bin/crsctl start
resource ora.asm -init
2012-09-01 18:23:03: Command output:
> CRS-2672: Attempting to start
'ora.asm' on 'gridnode1'
> CRS-2676: Start of 'ora.asm' on
'gridnode1' succeeded
>End Command output
2012-09-01 18:23:03: Start of resource "ora.asm" Succeeded
2012-09-01 18:23:03: Executing cmd:
/u01/app/11.2.0/grid/bin/crsctl start
resource ora.crsd -init
2012-09-01 18:23:04: Command output:
> CRS-2672: Attempting to start
'ora.crsd' on 'gridnode1'
> CRS-2676: Start of 'ora.crsd' on
'gridnode1' succeeded
>End Command output
2012-09-01 18:23:04: Start of resource "ora.crsd" Succeeded
2012-09-01 18:23:04: Executing cmd:
/u01/app/11.2.0/grid/bin/crsctl start
resource ora.evmd -init
2012-09-01 18:23:06: Command output:
> CRS-2672: Attempting to start
'ora.evmd' on 'gridnode1'
> CRS-2676: Start of 'ora.evmd' on
'gridnode1' succeeded
>End Command output
2012-09-01 18:23:06: Start of resource "ora.evmd" Succeeded
2012-09-01 18:23:06: Successfully started Oracle clusterware
stack
2012-09-01 18:23:06: Executing cmd: /u01/app/11.2.0/grid/bin/crsctl
stat resource
2012-09-01 18:23:08: Command output:
> CRS-4535: Cannot communicate
with Cluster Ready Services
> CRS-4000: Command Status failed,
or completed with errors.
>End Command output
2012-09-01 18:23:08: Waiting for Oracle CRSD and EVMD to
start
2012-09-01 18:23:13: Executing cmd: /u01/app/11.2.0/grid/bin/crsctl
stat resource
2012-09-01 18:23:13: Running /u01/app/11.2.0/grid/bin/crsctl check
cluster -n gridnode1
2012-09-01 18:23:13: Executing cmd: /u01/app/11.2.0/grid/bin/crsctl
check cluster -n gridnode1
2012-09-01 18:23:13: Command output:
> **************************************************************
> gridnode1:
> CRS-4535: Cannot communicate
with Cluster Ready Services
> CRS-4529: Cluster
Synchronization Services is online
> CRS-4534: Cannot communicate
with Event Manager
> **************************************************************
>End Command output
可以看到oracle GI启动的顺序是:
1、mDNSd
2、GPnPd
3、CSSD
cssd启动时候会依次启动:cssdmonitor、GPICd、diskmon、cssd
4、CTSSD
5、HAIP
6、ASM
7、CRSD
8、EVMD