Tools and Utilities - Expert Oracle RAC Performance Diagnostics and Tuning - page 170

ologgerd , like the name says, is a logger daemon process. It receives data from the osysmond
daemon and logs O/S metrics into a management db. One of the nodes in the cluster is the
designated master node and a second node in the cluster is chosen as the standby for the
master ologgerd . This ensures that statistics are collected on a continuous basis. Should the
master node fail, the standby ologgerd becomes the master and designates another server to
be the new standby.

•

• Berkeley DB (BDB)/Management DB: Data collected by the daemon process is retained in

a database for a period of 24 hours (default). This retention period can be increased to a
maximum period of 72 hours (259200 seconds) using the following command:

oclumon manage -bdb resize <size in number of seconds>
For example, oclumon manage -bdb resize 259200
As illustrated in Figure 6-21 , appropriate daemons are started on all nodes in the cluster. A single instance of
the database is created on the master node of the cluster. If the node fails, the database instance will be restarted on
another node.
CHM has a command line interface called oclumon . This utility allows querying the BDB to get node-specific
metrics for specific periods of time to view on a regular terminal. This is an interactive tool that can be used to get
node level or full cluster level information. For example, to get a detailed output about all nodes in the cluster, use

oclumon dumpnodeview -v allnodes - last "00:30:00"

CHM Statistics
Once CHM is installed and enabled, the daemons will start collecting data and saving it to the database where the
primary logger is currently residing. Statistics collected by CHM can be viewed using a command line interface with
oclumon utility.

[oracle@ssky1l4p1 ~]$ oclumon manage -get alllogger -details

Logger = ssky1l4p2
Nodes = ssky1l4p2,ssky1l4p1
[oracle@ssky1l4p1 ~]$ oclumon manage -get master

Master = ssky1l4p2

[oracle@ssky1l4p1 ~]$ oclumon manage -get repsize
CHM Repository Size = 136320

[oracle@ssky1l4p1 ~]$ oclumon manage -get reppath
CHM Repository Path = +GRID_DATA/_MGMTDB/DATAFILE/sysmgmtdata.260.846895815

[oracle@ssky1l4p1 ~]$ oclumon dumpnodeview -n ssky1l4p2
----------------------------------------

Node: ssky1l4p2 Clock: '14-05-24 23.32.53' SerialNo:11894

----------------------------------------