sp_logmon script does not display the correct backlog (4304441)

For Oracle 10g sp_logmon, displays the current redo log number instead of backlog values in the alerts.

When the Capture current redolog turns "0" (its random and will not show-up in the "show capture detail") the part of the script that calculates the redolog difference between the current Oracle redolog and the Shareplex redolog returns the current redolog number. This issue is observed with Oracle version 10g and in versions with redolog size > 2GB.

There are three workarounds to fix this issue.

Workaround 1 :

Upgrade Shareplex to the latest version of 7.5.x.

Workaround 2:

Add a condition to the script not to send notifications when the ${caplog} value is 0(which logically will not be 0 at any time).

Here is the part of the script that needs to be changed.

logdiff=`expr ${oralog} - ${caplog}`
if (test "$logdiff" -gt "$logthreshold")
then
echo "The SharePlex (port ${portnum}) Capture process has fallen behind in capturing" >logmon.err
echo "information from the redo log by ${logdiff} files, which exceeds the threshold (of" >>logmon.err

Workaround 3:

Use Shareplex Manager, it does all the shareplex monitoring and sends out notifications.

NOTE : Monitoring scripts are just a templates provided to the Customers to make changes as per their requirements. Customizations of the script extends beyond the scope of Support and is best handled by Quest Professional Services consultants who can help architect and build a solution that meets your needs.