I have a situation occurring that I am not sure is being caused by a problem with the database monitor agent or something in the way I have configured it.
The DB monitor is configured to perform a "select * from tb_SecurityLog" with a check interval of 15 seconds, overrun prevention delay of 1ms. This is a table that has a large volume of activity. The ID Field name is ID. If I set the "Last ID value" to some value in the middle of the database (confirmed by querying the DB directly) and start the agent, everything runs correctly for a while. It seems that when the Monitorware agent "catches up" and gets to the last record of the database one of two things are happening. One some occasions the Last ID value gets set to a very large value (for example last record is actually 3400000 but last ID value is set to 140000000) and then no additional records are read since the last ID value is greater than any record in the database. Other times when the agent gets to the end of the database, the last ID value gets set to -9661047 and this apparently causes the agent to begin reading at the beginning of the database again (my guess from looking at the output file). I can see the ID field incrementing in the output file and then it jumps to an earlier (at least 5000) record and begins working it's way back to the end.
Is this something being cause by the way I have it configured or some other issue?


