Insert trigger deadtime in the database

[lucrlom]

to calculate the total dead time of a run due to the busy trigger time can be called the function:
hax.trigger_data.get_trigger_data(run_number, 'dead_time_info')['busy'].sum()
But it takes so long time when you have to get it from several hundred file for the analysis.
It could be useful to insert it in the data base information for each run.

[JelleAalbers]

Relates to #430. If we decide to do this combined dead time calculation in the event builder I agree the result should be in the runs db. We may also want to do this in an external script; then still the result should be inserted in some database (but it would no longer be a pax issue... somehow I feel attracted to this option :-))

[JelleAalbers]

I think @coderdj calculates the dead time and inserts it under run_doc['quality']['daq']['extracted'], in one of the fields there. I don't see it in the most recent runs, but going back a while the information is there.

The multi-system (BUSY, HEV) deadtime calculation isn't implemented in the trigger, and it's easier to do offline anyway with other quality checks. However, it should be easy to tally the per-system deadtime and store it with the end-of-run info in the trigger.