was running into a similar issue when using process monitoring. We were building up more events than would expire in our timeframe, essentially meaning that we'e end up with the same msg you're getting