一次非常幸运的ORA-600 16703(tab$被清空)故障恢复---惜分飞

这次的ORA-600 16703的故障比较特殊,客户的一套rac运行了5年多没有重启,这次由于异常导致其中一个节点重启,然后触发了tab$被清空,异常节点启动报ORA-600 16703错误.朋友在故障之后,第一时间没有对在运行的节点进行重启(虽然也无法对外提供业务服务),使得恢复工作相对简单一些,恢复效果也是最完美的.这个是我在对于软件安装介质注入恶意脚本,300天之后重启触发tab$被清空的相关恢复case中,最完美的一次(以前遇到过一次客户是虚拟化环境通过cdp回退然后类似方法处理),凸显了这位朋友在故障发生之后对于问题的准确判断和果断的应对能力.
有朋友和我反馈,他们数据库突然报大量ORA-600错误,业务无法正常操作,我分析相关日志确认:节点2重启之后节点1开始报大量ORA-600错误,但是节点一直处于open状态

Fri Jul 25 15:28:53 2025 Decreasing number of real time LMS from 3 to 0 Fri Jul 25 15:29:18 2025 Reconfiguration started (old inc 13, new inc 15) List of instances:  1 2 (myinst: 1)  Global Resource Directory frozen  Communication channels reestablished  Master broadcasted resource hash value bitmaps  Non-local Process blocks cleaned out Fri Jul 25 15:29:18 2025  LMS 1: 0 GCS shadows cancelled, 0 closed, 0 Xw survived Fri Jul 25 15:29:18 2025  LMS 2: 0 GCS shadows cancelled, 0 closed, 0 Xw survived Fri Jul 25 15:29:18 2025  LMS 0: 0 GCS shadows cancelled, 0 closed, 0 Xw survived  Set master node info  Submitted all remote-enqueue requests  Dwn-cvts replayed, VALBLKs dubious  All grantable enqueues granted  Submitted all GCS remote-cache requests  Fix write in gcs resources Reconfiguration complete Fri Jul 25 15:29:20 2025 minact-scn: Master returning as live inst:2 has inc# mismatch instinc:0 cur:15 errcnt:0 Fri Jul 25 15:30:07 2025 Errors in file /u01/oracle/diag/rdbms/orcl/orcl1/trace/orcl1_ora_252634.trc  (incident=77234): ORA-00600: internal error code, arguments: [ktsircinfo_num1], [0], [0], [0], [],[],[],[],[],[],[],[] Incident details in: /u01/oracle/diag/rdbms/orcl/orcl1/incident/incdir_77234/orcl1_ora_252634_i77234.trc Fri Jul 25 15:30:18 2025 Use ADRCI or Support Workbench to package the incident. See Note 411.1 at My Oracle Support for error and packaging details. Fri Jul 25 15:30:19 2025 Sweep [inc][77234]: completed Sweep [inc2][77234]: completed Fri Jul 25 15:30:27 2025 Errors in <