你的agent日志显示2020-04-05 10:28:27.234开始stop database，但是你的alert最早的时间Wed Apr 08 11:26:09 2020，alert只能看到尝试了两次启动数据库最后都没成功，最好上传alert对应时间段的以及trace文件

Mos找到一篇文档，可以参考，但是不建议直接参照建议做，最好确定完整情况，再进行处理：

ORA-00600: [2702] during startup of one node (Doc ID 565940.1)

APPLIES TO:
Oracle Database - Enterprise Edition - Version 10.2.0.2 to 11.2.0.3 [Release 10.2 to 11.2]
Oracle Database - Enterprise Edition - Version 9.2.0.1 to 10.2.0.4 [Release 9.2 to 10.2]
Information in this document applies to any platform.

SYMPTOMS
The database fails to startup with ORA-00600:[2702] from one node.

The remaining nodes from the cluster start up just fine.

Tue Apr 29 12:27:33 2008
alter database open
Tue Apr 29 12:27:33 2008
This instance was first to open
Picked broadcast on commit scheme to generate SCNs
Tue Apr 29 12:27:33 2008
Errors in file /PATH/udump/FILE_NAME.trc:
ORA-00600: internal error code, arguments: [2702], [1], [9], [8], [], [], [], []
ORA-600 signalled during: alter database open…

CAUSE
Instance failed to startup as there was a ONLINE REDOLOG with status CURRENT and

ARCHIVED set to NO in v$log for thread that fails.

The same ONLINE REDOLOG has the lowest sequence within that thread.
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS

9 1 8673 52428800 2 NO CURRENT
10 1 8686 52428800 2 YES ACTIVE
11 1 8684 52428800 2 YES INACTIVE
12 1 8685 52428800 2 YES INACTIVE

SOLUTION
During startup, we look for to reuse the ONLINE REDOLOG with lowest sequence#.

That ONLINE REDOLOG Log must have be archived.
The V$LOG view shows that sequence# 8673 was not archived.

In order to fix it, you need to clear the REDOLOG with lower sequence#
shutdown all nodes
startup mount only one node
disable archivelog
open the database
disable the thread that has fails to startup
drop the group that contains the lowest sequence
recreate the REDOLOG group
renable the thread
shutdown the database
startup mount one node
re-enable archivelog
open the database

Oracle Database - Enterprise Edition - Version 10.2.0.2 to 11.2.0.3 [Release 10.2 to 11.2]???
这个问题处理只适用于11.2.0.3版本？
我这oracle集群是11.2.0.4版本数据库，linux是oracle linux server release 6.5版本，这种处理适用吗？

mos 说的 0.3 不一定就是到 0.3 为止，具体情况需要核对报错是否一样。

是的，JiekeXu说的对，都需要仔细核对报错是否相同，以及出现问题场景原因是否相同，不能盲目按照文档操作。