The Event Action Daemon stops processing new_events and triggers every hour.

The Event Action Daemon stops processing new_events and triggers every hour.

book

Article ID: KB0084997

calendar_today

Updated On:

Products Versions
TIBCO InConcert -
Not Applicable -

Description

Resolution:
Description: The Event Action Daemon in 3.7 may appear to stop processing NEW_EVENTS, TRIGGERS and
OLD_ACTIONS.  It will only be noticeable if you are using IcStatus to look at
the number of new_events or you have triggers that are delayed in firing.  This
will happen after the server has been up for the IC_EVENT_POLL_INTERVAL setting
and this is usually the default of 3600 seconds or 1 hour.

This will likely only happen with Oracle is running with an optimizer_mode of
'Chose' or 'Cost' AND there was a time when the number of NEW_EVENTS grew to the
thousands.  What happens is that Oracle makes some bad choices on a query that
the EA daemon makes at the IC_EVENT_POLL_INTERVAL and reads all blocks in the
NEW_EVENTS table up to the high water mark.

In 6.0 this should not be an issue because by default all operations are using RULE mode.

The solution is to either export and then import NEW_EVENTS table or to change

Issue/Introduction

The Event Action Daemon stops processing new_events and triggers every hour.

Environment

Product: TIBCO InConcert Version: 5.1.5, 5.1, 5.0, 4.6 OS: All --------------------