Find All Oracle Alert Logs

I know what you are thinking as I would have thought the same. What’s the big deal?

When there are dozen of RAC databases, it becomes cumbersome to search all alert logs one by one.

[grid@node1 ~]$ echo $ORACLE_BASE
/u01/app/grid
[grid@node1 ~]$ ls -l $ORACLE_BASE/diag/*/*/*/trace/alert_*.log
-rw-r-----. 1 grid oinstall 85909 Aug 21 13:30 /u01/app/grid/diag/asm/+asm/+ASM1/trace/alert_+ASM1.log
[grid@node1 ~]$ vi $ORACLE_BASE/diag/*/*/*/trace/alert_*.log

[grid@node2 ~]$ echo $ORACLE_BASE
/u01/app/grid
[grid@node2 ~]$ ls -l $ORACLE_BASE/diag/*/*/*/trace/alert_*.log
-rw-r-----. 1 grid oinstall 121172 Aug 21 13:30 /u01/app/grid/diag/asm/+asm/+ASM2/trace/alert_+ASM2.log
[grid@node2 ~]$ vi $ORACLE_BASE/diag/*/*/*/trace/alert_*.log


[oracle@node1 ~]$ echo $ORACLE_BASE
/u01/app/oracle
[oracle@node1 ~]$ ls -l $ORACLE_BASE/diag/*/*/*/trace/alert_*.log
-rw-r-----. 1 oracle asmadmin 0 Aug 21 13:40 /u01/app/oracle/diag/rdbms/db193h1/DB193H11/trace/alert_DB193H11.log
[oracle@node1 ~]$

[oracle@node2 ~]$ echo $ORACLE_BASE
/u01/app/oracle
[oracle@node2 ~]$ ls -l $ORACLE_BASE/diag/*/*/*/trace/alert_*.log
-rw-r-----. 1 oracle asmadmin 0 Aug 21 13:43 /u01/app/oracle/diag/rdbms/db193h1/DB193H12/trace/alert_DB193H12.log
[oracle@node2 ~]$

Advertisement

Cluster Interconnect Split Brain Testing (CSS_CRITICAL yes)

Simply, setting CSS_CRITICAL for the node will provide more weight to not shutdown node during splitbrain.

cluster_interconnect_split_brain_CSS_CRITICAL_yes

Cluster Interconnect Split Brain Testing

There was a dicusssion about memory causing split brain for a 2 nodes RAC; however, the log shows otherwise.

By no means, am I a RAC expert.

For testing purpose, ethernet for cluster_interconnect was shut down.

From ocssd.trc, it does look to be caused by cluster_interconnect when comparing errors from actual vs test.

Details from Google Doc below.

Cluster Interconnect Split Brain Testing

What do you think?