查看/var/log/secure日志发现rac节点之间相互有很多grid用户的ssh连接,类似
Aug 20 06:08:55 TEST-RAC-DB-02 sshd[51874]: Accepted publickey for grid from 192.168.1.123 port 38241 ssh2: RSA SHA256:xxxxxx
Aug 20 06:08:55 TEST-RAC-DB-02 sshd[51896]: Accepted publickey for grid from 192.168.1.123 port 38242 ssh2: RSA SHA256:xxxxxx
Aug 20 06:08:56 TEST-RAC-DB-02 sshd[51932]: Accepted publickey for grid from 192.168.1.123 port 38243 ssh2: RSA SHA256:xxxxxx
Aug 20 06:08:56 TEST-RAC-DB-02 sshd[51950]: Accepted publickey for grid from 192.168.1.123 port 38244 ssh2: RSA SHA256:xxxxxx
Aug 20 06:08:56 TEST-RAC-DB-02 sshd[51974]: Accepted publickey for grid from 192.168.1.123 port 38245 ssh2: RSA SHA256:xxxxxx
...
印象中rac只在安装和安装补丁时需要用到ssh,好奇为啥平时也有这么多ssh连接
经过一番苦逼的搜索,找到了发起ssh连接的原因 —— ora.cvu resource每隔一段时间通过ssh对集群进行监控检查,这是oracle的预期行为,并非异常现象
Frequent sshd Login by Grid User in Cluster Nodes by ora.cvu Resource (文档 ID 1943147.1)
查看cvu日志,根据官方文档,日志文件默认在 $ORACLE_BASE/crsdata/
host_name
/cvu
目录,文件名为cvutrace.log.0
The CVU trace files are created in the ORACLE_BASE/crsdata/
host_name
/cvu
directory by default. Oracle Database automatically rotates the log files and the most recently created log file has the name cvutrace.log.0
. You should remove unwanted log files or archive them to reclaim disk place if needed.
日志中可以看到类似如下信息:
[Worker 1] [ 2017-07-21 18:17:00.677 CST ] [UnixSystem.dorunRemoteExecCmd:3397] Final unix SSH command: /usr/bin/ssh -o FallBackToRsh=no -o PasswordAuthentication=no -o StrictHostKeyChecking=yes -o NumberOfPasswordPrompts=0 TEST-RAC-DB-01 -n /tmp/CVU_12.1.0.2.0_grid/exectask.sh -getver
关于CVU资源详细信息,参考
https://docs.oracle.com/database/121/CWADD/GUID-405E2CE7-D8B0-4D72-8E53-741DB6A8E919.htm#CWADD91382
https://docs.oracle.com/cd/B28359_01/rac.111/b28255/cvu.htm#BEHGEAEH