[Oracle 11g r2(11.2.0.4.0)]案例分析6-丢失磁盘心跳导致的集群重新配置

( 1 ) ocssd.log ( 10gR2 版本）

ERROR: Internal Error Information: Category: 1234 Operation: scls_block_writeLocation: fwrite faile Other: fwrite unable to write buffer Dep: 5

以上可知，VF无法访问。

<CSSD 2007-07-30 14:38:21.358 [1544] >TRACE: clssnmDiskPMT: stale disk (200012ms) (0//*********)CSSD 2007-07-30 14:38:21.358 [1544] >ERROR: clssnmDiskPMT: 1 of 1 voting disksunavailable (0／0／1)>

等待200s后， 对应的VF 被离线掉。

CSSD 2007-07-30 14: 38: 21. 358 [ 154 5] clssnmvDiskCheck: Aborting, 0 of 1 configuredvoting disks available, need 1CSSD 2007-07-30 14: 38: 21. 358 [ 154 5] ###################################CSSD 2007 07-30 14:38:21.358 [1546]clssscExit: CSSD aborting from thread clssnmvDiskPingMonitorThread

由于集群中只存在l块VF， 节点被重启。

( 2 ) ocssd.log ( 对于1 1 gR2 版本）

2010-08-13 18:31:19.787: [SKGFD] [4131736480]ERROR: -9(Error 27072, OS Error (Linux Error: 5: Input/output errorAdditional information: 4Additional information: 721425Additional information: -1)2010-08-13 18:31:19.787: [CSSD) [4131736480) (:CSSNM00060:)clssnmvReadB1ocks: readfailed at offset 529 of /*********

VF出现问题。

2010-08-13 18:34: 37. 423: [CSSD] [150477728] clssnmvDiskOpen: Opening /*********2010-08-13 18:34:37.423: [C L SF ] [150477728]openedhdl:Ox f4336530 for dev:/****：2010-08-13 18:34:37.429: [SKGFD] [150477728)ERROR: -9(Error 27072, OS Error (Linux Error: 5: Input/output errorAdditional information: 4Additional information: 720913Additional information: -1)2010-08-13 18:34:37.429: [CSSD] [150477728] (:CSSNM00060:)clssnmvReadB1ocks: readfailed at offset 17 of ／****2010-08-13 18:34:38.205: [CSSD] (4110736288] (:CSSNM00058:)clssnmvDiskCheck: No I/0 completions for 200880 ms for voting file /*********)

在等待200s后， 对应的VF 被离线掉。

2010-08-13 18:34:38.206: [CSSD] [4110736288] (:CSSN M00018:)clssnmv DiskCheck:Aborting, 0 of 1 configured voting disks ava工lable, need 12010-08-13 18: 34: 38. 206: [CSSD] [ 41107 36288] ###################################2010-08-13 18:34:38.206: [CSSD] [4110736288]clssscExit: CSSD aborting from thread clssnmvDiskPingMonitorThread2010-08-13 18: 34: 38. 20 6: [CSSD] [ 41107 36288] ###################################

由于集群中只存在1块VF， 节点被重启。对于11.2.0.2及以上版本的集群， 由于新特性reboot less restart的引人Oracle 会重新启动GI 而不会直接重启节点。