Suse11SP3系统平台Crsctl执行关闭时报错处理

Suse11SP3系统平台Crsctl执行关闭时报错处理Crsctl执行关闭时报错。jason1:/u01/app/11.2.0/grid/bin # ./crsctl stop crsCRS-2791: Starting shutdown of Oracle High AvailabilityServices-managed resources on 'jason1'CRS-2673: Attempting to stop 'ora.crsd' on 'jason1'CRS-2790: Starting shutdown of Cluster Ready Services-managedresources on 'jason1'CRS-2673: Attempting to stop 'ora.OCR.dg' on 'jason1'CRS-2673: Attempting to stop 'ora.registry.acfs' on 'jason1'CRS-2673: Attempting to stop 'ora.oracle.db' on 'jason1'CRS-2673: Attempting to stop 'ora.LISTENER_SCAN1.lsnr' on'jason1'CRS-2673: Attempting to stop 'ora.cvu' on 'jason1'CRS-2673: Attempting to stop 'ora.LISTENER.lsnr' on 'jason1'CRS-2673: Attempting to stop 'ora.oc4j' on 'jason1'CRS-2677: Stop of 'ora.LISTENER.lsnr' on 'jason1' succeededCRS-2673: Attempting to stop 'ora.jason1.vip' on 'jason1'CRS-2677: Stop of 'ora.cvu' on 'jason1' succeededCRS-2672: Attempting to start 'ora.cvu' on 'jason2'CRS-2677: Stop of 'ora.LISTENER_SCAN1.lsnr' on 'jason1'succeededCRS-2673: Attempting to stop 'ora.scan1.vip' on 'jason1'CRS-2677: Stop of 'ora.jason1.vip' on 'jason1' succeededCRS-2672: Attempting to start 'ora.jason1.vip' on 'jason2'CRS-2676: Start of 'ora.cvu' on 'jason2' succeededCRS-2677: Stop of 'ora.scan1.vip' on 'jason1' succeededCRS-2672: Attempting to start 'ora.scan1.vip' on 'jason2'CRS-2677: Stop of 'ora.oracle.db' on 'jason1' succeededCRS-2673: Attempting to stop 'ora.DATA.dg' on 'jason1'CRS-2673: Attempting to stop 'ora.FRA.dg' on 'jason1'CRS-2677: Stop of 'ora.registry.acfs' on 'jason1' succeededCRS-2676: Start of 'ora.jason1.vip' on 'jason2' succeededCRS-2676: Start of 'ora.scan1.vip' on 'jason2' succeededCRS-2672: Attempting to start 'ora.LISTENER_SCAN1.lsnr' on'jason2'CRS-2677: Stop of 'ora.OCR.dg' on 'jason1' succeededCRS-2677: Stop of 'ora.FRA.dg' on 'jason1' succeededCRS-2677: Stop of 'ora.DATA.dg' on 'jason1' succeededCRS-2673: Attempting to stop 'ora.asm' on 'jason1'CRS-2676: Start of 'ora.LISTENER_SCAN1.lsnr' on 'jason2'succeededCRS-2677: Stop of 'ora.asm' on 'jason1' succeededCRS-2677: Stop of 'ora.oc4j' on 'jason1' succeededCRS-2672: Attempting to start 'ora.oc4j' on 'jason2'CRS-2676: Start of 'ora.oc4j' on 'jason2' succeededCRS-2673: Attempting to stop 'ora.ons' on 'jason1'CRS-2677: Stop of 'ora.ons' on 'jason1' succeededCRS-2673: Attempting to stop 'ora.net1.network' on 'jason1'CRS-2677: Stop of 'ora.net1.network' on 'jason1' succeededCRS-2792: Shutdown of Cluster Ready Services-managed resourceson 'jason1' has completedCRS-2677: Stop of 'ora.crsd' on 'jason1' succeededCRS-2673: Attempting to stop 'ora.crf' on 'jason1'CRS-2673: Attempting to stop 'ora.ctssd' on 'jason1'CRS-2673: Attempting to stop 'ora.evmd' on 'jason1'CRS-2673: Attempting to stop 'ora.asm' on 'jason1'CRS-2673: Attempting to stop 'ora.mdnsd' on 'jason1'CRS-2673: Attempting to stop 'ora.drivers.acfs' on 'jason1'CRS-2677: Stop of 'ora.crf' on 'jason1' succeededCRS-2677: Stop of 'ora.evmd' on 'jason1' succeededCRS-2677: Stop of 'ora.mdnsd' on 'jason1' succeededCRS-2677: Stop of 'ora.ctssd' on 'jason1' succeededCRS-5022: Stop of resource "ora.drivers.acfs" failed:current state is "UNKNOWN"CRS-2675: Stop of 'ora.drivers.acfs' on 'jason1' failedCRS-2677: Stop of 'ora.asm' on 'jason1' succeededCRS-2673: Attempting to stop 'ora.cluster_interconnect.haip' on'jason1'CRS-2677: Stop of 'ora.cluster_interconnect.haip' on 'jason1'succeededCRS-2673: Attempting to stop 'ora.cssd' on 'jason1'CRS-2677: Stop of 'ora.cssd' on 'jason1' succeededCRS-2673: Attempting to stop 'ora.gipcd' on 'jason1'CRS-2677: Stop of 'ora.gipcd' on 'jason1' succeededCRS-2673: Attempting to stop 'ora.gpnpd' on 'jason1'CRS-2677: Stop of 'ora.gpnpd' on 'jason1' succeededCRS-2799: Failed toshut down resource 'ora.drivers.acfs' on 'jason1'CRS-2795: Shutdownof Oracle High Availability Services-managed resources on 'jason1' has failedCRS-4687: Shutdowncommand has completed with errors.CRS-4000: CommandStop failed, or completed with errors.jason1:/u01/app/11.2.0/grid/bin # 查看crs日志发现每次关闭时否有如下日志存在CRS-10001:07-Jan-16 23:23 ACFS-9118: oracleacfs.ko driver in use - cannot unload证明每次在关闭资源ora.drivers.acfs时，还有程序在使用该模块导致关闭失败。查看MOS内容如下：

SLES: 11.2.0.3 "crsctl stop crs" Fails to Stop ora.drivers.acfs With CRS-2675 (文档 ID 1417294.1)

In this Document APPLIES TO: Oracle Database - Enterprise Edition - Version 11.2.0.3 and later SUSE \ UnitedLinux x86-64 SYMPTOMS 11.2.0.3 Grid Infrastructure on Suse Linux, "crsctl stop crs" or "crsctl stop crs -f" fails: "crsctl stop crs" or "crsctl stop crs -f" output CRS-2675: Stop of 'ora.drivers.acfs' on 'racnode1' failed .. CRS-2799: Failed to shut down resource 'ora.drivers.acfs' on 'racnode1' CRS-2795: Shutdown of Oracle High Availability Services-managed resources on 'racnode1' has failed CRS-4687: Shutdown command has completed with errors. CRS-4000: Command Stop failed, or completed with errors. $GRID_HOME/log/<node>/alert<node>.log [client(4936)]CRS-10001:01-Feb-12 12:35 ACFS-9290: Waiting for ASM to shutdown. [client(5098)]CRS-10001:01-Feb-12 12:35 ACFS-9118: oracleacfs.ko driver in use - cannot unload. [client(5117)]CRS-10001:01-Feb-12 12:35 ACFS-9118: oracleacfs.ko driver in use - cannot unload. /log/racnode1/agent/ohasd/orarootagent_root/orarootagent_root.log" "$GRID_HOME/bin/acfsload stop" output .. ACFS-9119: oracleacfs driver failed to unload. $GRID_HOME/log/<node>/agent/ohasd/orarootagent_root/orarootagent_root.log 2012-02-15 09:44:43.219: [ AGFW][2097063696] {0:0:36062} Agent received the message: RESOURCE_STOP[ora.drivers.acfs 1 1] ID 4099:217493 2012-02-15 09:44:43.219: [ AGFW][2097063696] {0:0:36062} Preparing STOP command for: ora.drivers.acfs 1 1 2012-02-15 09:44:43.219: [ AGFW][2097063696] {0:0:36062} ora.drivers.acfs 1 1 state changed from: ONLINE to: STOPPING 2012-02-15 09:44:43.220: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] (:CLSN00108:) clsn_agent::stop { 2012-02-15 09:44:43.220: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] Arg Value = stop 2012-02-15 09:44:43.220: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] getOracleHomeAttrib: oracle_home = /ocw/grid 2012-02-15 09:44:43.220: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] Utils::getCrsHome crsHome /ocw/grid .. 2012-02-15 09:44:43.220: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] Adding Environment Variables _ORA_AGENT_ACTION=TRUE 2012-02-15 09:44:43.220: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] Adding Environment Variables __IS_HASD_AGENT= 2012-02-15 09:44:43.220: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] Utils:execCmd action = 2 flags = 6 ohome = (null) cmdname = acfsload. 2012-02-15 09:44:43.220: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] getOracleHomeAttrib: oracle_home = /ocw/grid 2012-02-15 09:44:43.321: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] (:CLSN00010:)ACFS-9290: Waiting for ASM to shutdown. 2012-02-15 09:44:43.321: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] (:CLSN00010:) .. 2012-02-15 09:44:53.433: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] (:CLSN00010:)ACFS-9118: oracleacfs.ko driver in use - cannot unload. 2012-02-15 09:44:53.433: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] (:CLSN00010:) 2012-02-15 09:44:53.433: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] execCmd ret = 0 2012-02-15 09:44:53.434: [ora.drivers.acfs][2094962448] {0:0:36062} [stop] (:CLSN00108:) clsn_agent::stop } 2012-02-15 09:44:53.434: [ AGFW][2094962448] {0:0:36062} Command: stop for resource: ora.drivers.acfs 1 1 completed with status: SUCCESS CAUSE The issue was investigated in bug 13726093 and it's closed as OS issue SOLUTION Engage OS vendor to disable software that scans ACFS devices. The temporary workaround is to disable the resource if ACFS is not needed. To disable: # $GRID_HOME/bin/acfsroot disable Once it's disabled and node rebooted, it's status will be ora.registry.acfs ONLINE OFFLINE racnode1 Note: It's known that multipath daemon may open /dev/ofsctl: lsof /dev/ofsctl COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME multipath 5356 root 58r BLK 251,0 0t0 41683748 /dev/ofsctl Adding ofsctl to multipath blacklist solves the issue. for findability: rootcrs.pl -unlock ; roothas.pl -unlock REFERENCES BUG:13606669 - CRSCTL STOP CLUSTERWARE FAILS TO STOP, ACFS-9118 BUG:13613644 - CRSCTL STOP CRS IS NOT STOPPING THE ORA.DRIVERS.ACFS RESOURCE 11.2.0.3 BUG:13726093 - CRSCTL STOP CRS FAILS TO STOP ACFS BUG:13736590 - CRS-2799: FAILED TO SHUT DOWN RESOURCE 'ORA.DRIVERS.ACFS' BUG:13810374 - FAILS TO STOP ACFS RESOURCE IN 11.2.0.3

因本实例中使用multipath绑定硬盘作为ASM磁盘，故multipath服务打开/dev/ofsctl设备，所以导致ora.drivers.acfs失败，最终整个关闭命令完成带有错误。根据提示在multipath配置文件中设置blacklist{ devnode "^sda" devnode "ofsctl"}设置后，再次重试关闭命令问题解决。