Monitor tag value changes

Having the possibility to monitor tag value changes as a batch query would be extremely useful.

For example instead of using the deadman function you could compare the server tag values for the last 10 minutes to the values 60+10 minutes ago to see if any servers stopped sending in data. What this also would allow us to do is to monitor for new servers sending in data.

Start by reading Kapacitor's existing deadman function and batch-query handling to understand how time ranges and tag values are represented. Define how a query would compare tag values across the two periods and how it would identify missing or newly reporting servers; done means the behavior is specified and covered by appropriate query tests.

Written by the indexing model from the issue text.