The Common Agent Container cacao is a common Java container for JDMX/JMX management and handles the interactions between controllers and agents. All events are recorded in the cacao log files. Any event above the level of INFO is also logged in the syslog. You can view the contents of the current log file using the UI or by viewing the contents of the following files:
On an Oracle Solaris Enterprise Controller: /var/cacao/instances/oem-ec/logs/cacao.n
On a Linux Enterprise Controller: /var/opt/sun/cacao/instances/oem-ec/logs/cacao.n
On each Oracle Solaris Proxy Controller: /var/cacao/instances/scn-proxy/logs/cacao.n
On each Linux Proxy Controller: /var/opt/sun/cacao/instances/scn-proxy/logs/cacao.n
On each Oracle Solaris agent: /var/cacao/instances/scn-agent/logs/cacao.n
On each Oracle Linux agent: /var/opt/sun/cacao/instances/scn-agent/logs/cacao.n
The maximum file size is 1 MB. When the limit is reached, the current log file is closed and a new one created. The default number of log files is three. You can change the number of log files that are retained, using the Common Agent Container's management utility, cacaoadm.
To view the current number of log files maintained for the Enterprise Controller, issue the following command on the system where the Enterprise software is running:
# cacaoadm get-param log-file-count -i oem-ec
log-file-count=3
To view the number of log files maintained for a Proxy Controller, issue the following command on the system were the proxy controller software is running:
# cacaoadm get-param log-file-count -i scn-proxy
log-file-count=3
To view the number of log files maintained for an agent use the following command on the system were the agent is running:
# cacaoadm get-param log-file-count -i scn-agent
log-file-count=3
To Change the Number of Log Files for an Enterprise Controller or Proxy Controller
To change the number of log files on the Enterprise Controller, a Proxy Controller, or both:
Verify that there are no active jobs.
Stop the Common Agent Container service on the Enterprise Controller.
On Oracle Solaris:
# /opt/SUNWxvmoc/bin/satadm stop -w -v
On Linux:
# /opt/sun/xvmoc/bin/satadm stop -w -v
Stop the Common Agent Container service on a Proxy Controller:
On Oracle Solaris:
# /opt/SUNWxvmoc/bin/proxyadm stop -w -v
On Linux:
# /opt/sun/xvmoc/bin/proxyadm stop -w -v
Specify the maximum number of log files to be retained in addition to the current log file. In the following example, the count of 10 specifies that nine log files of previous events are retained in addition to the log file for current events. On the Enterprise Controller:
# cacaoadm set-param log-file-count=10 -i default
On a Proxy Controller:
# cacaoadm set-param log-file-count=10 -i scn-proxy
Start the Enterprise Controller.
On Oracle Solaris:
# /opt/SUNWxvmoc/bin/satadm start -w -v
On Linux:
# /opt/sun/xvmoc/bin/satadm start -w -v
Verify that the Enterprise Controller has been restarted completely before attempting other operations. For example, if you have stopped both the Enterprise Controller and a Proxy Controller, wait for the Enterprise Controller to restart before restarting each Proxy Controller.
On an Oracle Solaris Proxy Controller:
# /opt/SUNWxvmoc/bin/proxyadm start -w -v
On a Linux Proxy Controller:
# /opt/sun/xvmoc/bin/proxyadm start -w -v
Verify that all controllers have restarted completely before attempting other operations.
To Change the Number of Log Files for an Agent
To change the number of log files on an agent:
Verify that there are no active jobs.
Stop the Common Agent Container service on a the agent:
On Oracle Solaris:
# /opt/SUNWxvmoc/bin/agentadm stop -v
On Linux:
# /opt/sun/xvmoc/bin/agentadm stop -v
Specify the maximum number of log files to be retained in addition to the current log file. In the following example, the count of 10 specifies that nine log files of previous events are retained in addition to the log file for current events.
# cacaoadm set-param log-file-count=10 -i scn-agent
Start the agent:
On Oracle Solaris:
# /opt/SUNWxvmoc/bin/agentadm start -v
On Linux:
# /opt/sun/xvmoc/bin/agentadm start -v
Verify that all controllers have restarted completely before attempting other operations.