Thanks for the response. For #3, any thoughts on more granular configurable settings for the alert types, as in the example I gave? Some alerts might be more important than others, so we could let other expire faster while more critical errors stay up longer in the list