News

  1. Home
  2. All posts
All posts, Exa DATA, Oracle by Team Ajara

dcli on exadata, trace file generation is huge

drwXIUGAJHGD

dcli command example on exadata : one Issue with trace file generation is huge

In this post ,we will provide how to use the dcli command with one example with issue.

Overview of the issue:

One DB was configured on four node RAC in the full Rack exadata machine(8 DB Nodes).

 

Root cause

Due to configuration of KRB evenet in the database ,huge trace files got generated.

There are huge Number of trace files are getting generated for CRMD database with “KRB” Error.
After checking, due to below parameter, huge number of trace files are getting generated.
==>ALTER SYSTEM SET event=’10720 trace name context forever, level 16384′,’logon trace name krb_options level 26′ SCOPE=SPFILE;
As per the Error, looks like it was set as part of RMAN Duplicate or some other Restoration.

Verify the size of trace directory on each node

[oracle@Node01 ~]$ dcli -l oracle -g dbs_group_CRMD ” du -sh /u01/app/oracle/diag/rdbms/CRMD/CRMD*/trace/”
Node01: 99G /u01/app/oracle/diag/rdbms/CRMD/CRMD1/trace/
Node02: 89G /u01/app/oracle/diag/rdbms/CRMD/CRMD2/trace/
Node03: 96G /u01/app/oracle/diag/rdbms/CRMD/CRMD3/trace/
Node04: 151G /u01/app/oracle/diag/rdbms/CRMD/CRMD4/trace/ —-> Huge size before trim

Verify the alert log of the 4th node

[oracle@Node04 trace]$ cat alert_CRMD4.log | grep -i krb

ALTER SYSTEM SET event=’10720 trace name context forever, level 16384′,’logon trace name krb_options level 26′ SCOPE=SPFILE;
event = “logon trace name krb_options level 26”

[oracle@Node04 trace]$

 

Solution

Trim the trace directory on 4th node

compress all trace files older than 1 day.

Verify the size of trace directory on each node

Now The size on 4th node is small, as we have compressed all trace files older than 1 day.

[oracle@Node01 ~]$ dcli -l oracle -g dbs_group_CRMD ” du -sh /u01/app/oracle/diag/rdbms/CRMD/CRMD*/trace/”
Node01: 99G /u01/app/oracle/diag/rdbms/CRMD/CRMD1/trace/
Node02: 89G /u01/app/oracle/diag/rdbms/CRMD/CRMD2/trace/
Node03: 96G /u01/app/oracle/diag/rdbms/CRMD/CRMD3/trace/
Node04: 48G /u01/app/oracle/diag/rdbms/CRMD/CRMD4/trace/
[oracle@Node01 ~]$ dcli -l oracle -g dbs_group_CRMD “ls -ltrh /u01/app/oracle/diag/rdbms/CRMD/CRMD*/trace/| wc -l”
Node01: 404813
Node02: 503279
Node03: 464711
Node04: 459649
[oracle@Node01 ~]$

Related posts