We are being flooded by internal events for items, to the point of DB lockup.

Steps to reproduce:
This was probably occurring since our 3.4 conversion, but the events are for disabled hosts.

Result:
SQL query for the last hour:
set @unixnow = unix_timestamp();
set @friendlynow = from_unixtime(@unixnow);
set @friendlythen = date_sub(@friendlynow, interval 1 hour);
set @unixthen = unix_timestamp(@friendlythen);

select events.source,
events.object,
hosts.name,
hosts.status,
count
from events
inner join items
on events.objectid = items.itemid
inner join hosts
on items.hostid = hosts.hostid
where clock > @unixthen and events.source = '3' and events.object = '0'
group by events.source, events.object, hosts.name, hosts.status

Sample output of this query:
source object name status count
3 0 Legacy-c2960s004-South-2300 1 11
3 0 Legacy-c2960s004-Telecom-110 1 28
3 0 Legacy-c2960s005-UniversityHall-124 1 839
3 0 Legacy-c2960s005-WalterWilliams-S005A 1 80
3 0 Legacy-c2960s006-UniversityHall-124 1 55
3 0 Legacy-c2960s006-WalterWilliams-246C 1 80
3 0 Legacy-c2960s016-VetDiagLab-D130 1 15
3 0 Legacy-c2960s017-CAHousing-3002 1 736
3 0 Legacy-c2960s019-Schurz-8 1 105
3 0 Legacy-c2960s022-Hatch-9 1 958
3 0 Legacy-c2960s026-EBN-110 1 864
3 0 Legacy-c2960s033-Agriculture-1-5 1 167
3 0 Legacy-c2960s037-MATC-Weight 1 120
3 0 Legacy-c2960s042-ASRC-171 1 2694