Methods to shut down oracle rac: 1. Stop the listening service of each node; 2. Shut down the database instance; 3. Shut down the ASM instance; 4. Use the crs_stat command to check whether the service status of the RAC node is normal; 5. Stop CRS cluster software is enough.

The operating environment of this tutorial: Windows 7 system, Oracle 11g version, Dell G3 computer.

How to start and shut down Oracle RAC correctly?

If you want to shut down the Oracle cluster database system, the simple and crude way is to execute the OS command on the operating system: shutdown -h now;

This method is more violent and can shut down the database system. However, some Oracle services may not be shut down properly. After restarting the OS, many services will be in the UNKNOW state.

So how to shut down and start RAC correctly according to the Oracle working mechanism?

##RAC Reasonable order for shutdown: (Note: Exit all client connections, such as SQLplus...)

Stop listening on each node Service-->>Close database (instance)-->>Close ASM instance-->>Close node cluster service-->>Close server

1. Stop the monitoring service of each node

Check the listening status: ##

srvctl?status?listener?-n?node1

##Close listening:

[oracle@node1?~]$?srvctl?stop?listener?-n?node1
[oracle@node2?~]$?srvctl?stop?listener?-n?node2

You can use the above command to stop the corresponding monitoring on two nodes, or you can use the following command to execute on one node to stop All monitoring.

[oracle@node1?~]$?srvctl?stop?listener

Check the listening status again:

2. Close the database instance

For example, the database instance name found is: DEVDB

(2) Use the srvctl (service control) command to shut down the cluster services in sequence (note: exit the sqlplus client above, otherwise the database cannot be shut down normally)

srvctl status database -d DEVDB （數(shù)據(jù)庫實(shí)例名根據(jù)實(shí)際填寫，并且區(qū)分大小寫)

b Close the database instance:

srvctl stop database -d DEVDB

The above command can only be used in one You can execute it on the node, or you can execute the following commands separately on two nodes to achieve the same function.

srvctl stop instance -d DEVDB -i devdb1  （填寫自實(shí)際的INSTANCE_NAME）
srvctl stop instance -d DEVDB -i devdb2

C Check the shutdown status: srvctl status database -d DEVDB

3.關(guān)閉ASM實(shí)例

查看ASM運(yùn)行狀態(tài)：srvctl status asm -n node1

停止ASM實(shí)例：

srvctl stop asm -n node1

4.關(guān)閉RAC(CRS)集群服務(wù)

(1)使用crs_stat 命令查詢RAC節(jié)點(diǎn)的服務(wù)狀態(tài)是否正常 (root或者grid用戶)

crs_stat -t -v

(2)停止CRS集群服務(wù)

停止兩個(gè)節(jié)點(diǎn)的服務(wù)：

srvctl stop nodeapps -n node1
srvctl stop nodeapps -n node2

****

(3)停止CRS集群軟件

[root@node1 ~]# find / -name crsctl 

[root@node1 ~]# /u01/grid/11.2.0.4/bin/crsctl stop crs   （分別在兩個(gè)節(jié)點(diǎn)的root用戶下執(zhí)行）

CRS-2791: 正在啟動(dòng)用于關(guān)閉 'rac1' 上 Oracle High Availability Services 管理的資源的操作
CRS-2673: 嘗試停止 'ora.crsd' (在 'rac1' 上)
CRS-2790: 正在啟動(dòng)關(guān)閉 'rac1' 上集群就緒服務(wù)管理的資源的操作
CRS-2673: 嘗試停止 'ora.asm' (在 'rac1' 上)
CRS-2677: 成功停止 'ora.asm' (在 'rac1' 上)
CRS-2792: 關(guān)閉 'rac1' 上集群就緒服務(wù)管理的資源的操作已完成
CRS-2677: 成功停止 'ora.crsd' (在 'rac1' 上)
CRS-2673: 嘗試停止 'ora.cssdmonitor' (在 'rac1' 上)
CRS-2673: 嘗試停止 'ora.ctssd' (在 'rac1' 上)
CRS-2673: 嘗試停止 'ora.evmd' (在 'rac1' 上)
CRS-2673: 嘗試停止 'ora.asm' (在 'rac1' 上)
CRS-2673: 嘗試停止 'ora.mdnsd' (在 'rac1' 上)
CRS-2677: 成功停止 'ora.cssdmonitor' (在 'rac1' 上)
CRS-2677: 成功停止 'ora.evmd' (在 'rac1' 上)
CRS-2677: 成功停止 'ora.mdnsd' (在 'rac1' 上)
CRS-2677: 成功停止 'ora.ctssd' (在 'rac1' 上)
CRS-2677: 成功停止 'ora.asm' (在 'rac1' 上)
CRS-2673: 嘗試停止 'ora.cssd' (在 'rac1' 上)
CRS-2677: 成功停止 'ora.cssd' (在 'rac1' 上)
CRS-2673: 嘗試停止 'ora.gpnpd' (在 'rac1' 上)
CRS-2673: 嘗試停止 'ora.diskmon' (在 'rac1' 上)
CRS-2677: 成功停止 'ora.gpnpd' (在 'rac1' 上)
CRS-2673: 嘗試停止 'ora.gipcd' (在 'rac1' 上)
CRS-2677: 成功停止 'ora.gipcd' (在 'rac1' 上)
CRS-2677: 成功停止 'ora.diskmon' (在 'rac1' 上)
CRS-2793: 關(guān)閉 'rac1' 上 Oracle High Availability Services 管理的資源的操作已完成
CRS-4133: Oracle High Availability Services has been stopped.

注意：停止數(shù)據(jù)庫集群軟件，在一個(gè)節(jié)點(diǎn)執(zhí)行即可，在節(jié)點(diǎn)2就不用執(zhí)行了。

(4)再次查看兩個(gè)節(jié)點(diǎn)的CRS服務(wù)狀態(tài)：

[grid@node1 ~]$ crs_stat -t -v 
CRS-0184: Cannot communicate with the CRS daemon. （服務(wù)已經(jīng)關(guān)閉）
[grid@node2 ~]$ crs_stat -t -v
CRS-0184: Cannot communicate with the CRS daemon. （服務(wù)已經(jīng)關(guān)閉）

當(dāng)上述操作都執(zhí)行完成，并且正常關(guān)閉后，就可以將服務(wù)器的關(guān)閉了。

================

啟動(dòng)的順序和關(guān)閉相反

啟動(dòng)服務(wù)器-->>啟動(dòng)服務(wù)-->>啟動(dòng)ASM-->>啟動(dòng)實(shí)例-->>啟動(dòng)監(jiān)聽

這里我們需要注意；Oracle RAC集群軟件有一個(gè)屬性可以設(shè)置，就是當(dāng)服務(wù)器操作系統(tǒng)重啟后，集群服務(wù)是否會(huì)隨著操作系統(tǒng)一塊啟動(dòng)；上邊啟動(dòng)的順序步驟是在：當(dāng)禁用該自啟動(dòng)屬性的情況下的操作順序；關(guān)于該屬性的查看與設(shè)置，詳見下一篇穩(wěn)文章：關(guān)于集群自啟動(dòng)的查看。

(1)首先啟動(dòng)服務(wù)器

(2)啟動(dòng)CRS軟件：

[root@node1 ~]# find / -name crsctl 
[root@node1 ~]# /u01/grid/11.2.0.4/bin/crsctl start crs

查看服務(wù)：

srvctl status nodeapps -n node1
srvctl status nodeapps -n node2
crsctl status res – -t

(3)啟動(dòng)ASM

查看ASM狀態(tài)

srvctl status asm -n node1

啟動(dòng)ASM

srvctl start asm -n node1
srvctl start asm -n node2

(4)啟動(dòng)實(shí)例

srvctl status database -d node1
srvctl start database -d node2

(5)啟動(dòng)監(jiān)聽

srvctl status listener   （RAC環(huán)境）
srvctl start listener  
或者 lsnrctl start

===============================

重啟=關(guān)閉+啟動(dòng)

從上述各步可以看出，關(guān)閉和開啟都是分步執(zhí)行的，這樣操作還有一個(gè)好處，可以幫助我們發(fā)現(xiàn)定位問題。

推薦教程：《Oracle使用教程》

The above is the detailed content of How to close oracle rac. For more information, please follow other related articles on the PHP Chinese website!