HACMP5.1裝好oralce后的問題

★jasson★

smitty clstop停止時報warning如下：\r\nexist.\r\n# tail -f /tmp/hacmp.out\r\n:config_too_long[95] [ -x /usr/lpp/ssp/bin/spget_syspar ]\r\nWARNING: Cluster cls22 has been running recovery program \'/usr/es/sbin/cluster/events/node_down.rp\' for 360 seconds. Please check cluster status.\r\nWARNING: Cluster cls22 has been running recovery program \'/usr/es/sbin/cluster/events/node_down.rp\' for 390 seconds. Please check cluster status.\r\nWARNING: Cluster cls22 has been running recovery program \'/usr/es/sbin/cluster/events/node_down.rp\' for 420 seconds. Please check cluster status.\r\nWARNING: Cluster cls22 has been running recovery program \'/usr/es/sbin/cluster/events/node_down.rp\' for 450 seconds. Please check cluster status.\r\nWARNING: Cluster cls22 has been running recovery program \'/usr/es/sbin/cluster/events/node_down.rp\' for 480 seconds. Please check cluster status.\r\nWARNING: Cluster cls22 has been running recovery program \'/usr/es/sbin/cluster/events/node_down.rp\' for 540 seconds. Please check cluster status.\r\nWARNING: Cluster cls22 has been running recovery program \'/usr/es/sbin/cluster/events/node_down.rp\' for 600 seconds. Please check cluster status.\r\nWARNING: Cluster cls22 has been running recovery program \'/usr/es/sbin/cluster/events/node_down.rp\' for 660 seconds. Please check cluster status.\r\nWARNING: Cluster cls22 has been running recovery program \'/usr/es/sbin/cluster/events/node_down.rp\' for 720 seconds. Please check cluster status.\r\n\r\n\r\n停止不了\r\n誰知道問題所在？

yanbing

要看hacmp.out的故障內(nèi)容。。。\r\n\r\n你貼出來的這個是最后的超時提示（是故障后的結(jié)果）

★jasson★

下個禮拜去現(xiàn)場再貼hacmp.out出來\r\n還有一個現(xiàn)象就是，我的HACMP是并發(fā)版，共享卷組建了兩個，兩個vg再同一個資源組里面\r\n再down HA的時候，其中有一臺經(jīng)常只有一個vg down下來，還有一個vg始終down不下來、

bobqi

你可以試試打上oracle的最新補丁，\r\n現(xiàn)在是oracle9i的話可以到9.2.0.6

RSTSG

等你的hacmp.out 。\r\n\r\n把那臺出現(xiàn)超時的node的hacmp。out貼上來就可以了！\r\n\r\n根據(jù)你后面的描述，應該就是vg釋放不了的那個node有超時現(xiàn)象

RSTSG

看起來像是oracle沒有能夠停止下來�。�！\r\n\r\n如果方便的話，請把下面這個腳本的內(nèi)容貼出來看看。。\r\n\r\n /etc/qwaresh/stopedzsh 。\r\n\r\n原因應該就在這里

leijunxiang

我在p690上的hacmp也有這樣的情況，老是有時間的告警。但是每次都會成功切換。大概是劃分的表空間比較多的原因吧！

★jasson★

腳本如下：\r\nbanner \"stop!!!\" >;>; /tmp/hacmp.out\r\nsu - oracle \" -c lsnrctl stop\"\r\nsu - oracle \" -c dbshut\"

RSTSG

問題就在腳本上。。。\r\n\r\n你用su - oracle -c dbshut來停止數(shù)據(jù)庫。。。\r\n\r\n但是dbshut里面的shutdown是沒有參數(shù)的。也就是說一旦有一個客戶端還連接在你的oracle上，你的oracle就停不下來。\r\n\r\n建議解決方案：\r\n1。將dbshut里面的shutdown都帶上immediate參數(shù)；\r\n\r\n2�；蛘咧苯訉憇qlplus /nolog的腳本內(nèi)容，也是用immediate參數(shù)；\r\n\r\n3。如果你的客戶端實在是很多，應該考慮abort方式。

★jasson★

RSTSG\r\n麻煩你，怎么帶immediate參數(shù)？\r\n是直接寫dbshut immediate還是shutdown immediate，還是修改dbshut，添加參數(shù)？？