Weekly sweep seems to run all day on a very large site, or competes with content indexing

What you are seeing

The weekly sweep seems to run all day on a very large site, or competes with content indexing.


#Why it happens

It runs around 16 count queries per space paced at about five requests a second, with an 800-second budget per invocation and a checkpoint every 250 spaces, and it shares the app's single background queue with content indexing. Rate limiting pauses and resumes it.


#What to do

Let it finish rather than pressing Refresh repeatedly, since Refresh restarts the whole sweep.


#Settings involved

Setting

Where

What it does

Not updated in (Stale content)

Rules → Stale content → Not updated in

-

Refresh (Dashboard)

Capable Quality → Dashboard → Refresh

-


#Limits that apply here

Limit

Value

What happens

Trend history retention

60 weekly points and 36 monthly points;

-

Cross-space data freshness

Up to a week old

-

Weekly sweep pacing and budget

About 16 count queries per space, paced

-


#A few things that catch people out

  • Only the last 60 weekly and 36 monthly trend points are kept; older points are dropped.



Found it? Good.