Using High Availability Monitors (June 2003)

Troubleshooting HA Monitors
Database Monitor Troubleshooting for Oracle
Appendix B86
Database Monitor Troubleshooting for Oracle
For Oracle MIB monitors, review the following troubleshooting hints.
• If MIB resource classes under rdbms continue to be unavailable,
there might be a problem with the Oracle SNMP daemons. For
Oracle, these are ora_naaagt, master_peer, dbsnmp, or tnslsnr.
Try using the following commands to stop and restart Oracle SNMP:
su oracle -c “cd $ORACLE_HOME/bin; ./lsnrctl stop”
(cd $ORACLE_HOME/bin; ./lsnrctl dbsnmp_stop)
(cd $ORACLE_HOME/network/snmp/peer; ./start_peer -k -a)
su oracle -c "cd $ORACLE_HOME/bin; ./lsnrctl start"
(cd $ORACLE_HOME/bin; ./lsnrctl dbsnmp_start)
You might have to wait up to 60 secondsbefore trying toaccess rdbms
resources again.
•If rdbms resource classes continue to be unavailable, be sure that one
of the get-community and set-community names in
/etc/SnmpAgent.d/snmpd.conf match the -c option on each of the
MONITOR statements in the mibmond.dict and rdbmsmond.dict files
under /etc/opt/resmon/dictionary.
If changes are needed to dictionary files, stop any MIB monitors
that are already running.
If changes are needed to snmpd.conf, stop and restart HP SNMP,
following the procedure above.
NOTE The -c option is not required if public is one of the get-community
or set-community names.
•If rdbms resource classes continue to be unavailable, be sure that the
HP_NAA_GET_COMMUNITY name in the start_peer script matches the
COMMUNITY name in the CONFIG.master file. Both files are under
$ORACLE_HOME/network/snmp/peer.
If any changes are needed, it will be necessary to stop and restart
Oracle SNMP, following the procedure above.