Is there known bug in DB Connect causing scheduled jobs to run twice.. We had an incident over the weekend and messed up the report showing doubled events
Not that I know... but are you using DB Dump or DB Tail?
If you're using DB Tail, it stores the last value at $SPLUNK_HOME/var/lib/splunk/persistentstorage/dbx, where each DB Tail configuration has its own directory and it saves the value for the "tailing" process insice state.xml.
If the process failed in the middle or something happen where this file wasn't updated, the next run will insert duplicated entries.