5.6.4. FMS with DBO 5.6.4.311 Agent monitoring a 2 node cluster (Oracle 11.1.07).
Whenever the SELECT * from Q_Alert_Log..... (query to parse the alert logs) is ran is run by the DBO agent the CPU usage on the DBO Instance server reaches 100% and stays there
Agent logs we see that the collection works as expected in most of the time, but if there is a restart to one of the RAC nodes (from Oracle alert logs), a socket time error appears.
the Agent executes JDBC connection which reflects the timeout.
This leads us to belive that there is something in the connectivity (on the db server side/instance) that causes this scenario.
Change the socket timeout from the default 15 minutes to 90 seconds (Agent Status dashboard | select agent | Edit Properties | Monitored Connection Profile | DB read timeout while reading from the sockets), for each of the RAC nodes.
© ALL RIGHTS RESERVED. Feedback Terms of Use Privacy Cookie Preference Center