Disable flattening / Make it on-demand
Closed this issue · 2 comments
jpmckinney commented
Like cove-oc4ids, flattening uses an inordinate amount of RAM. Since it's infrequently used (though more than on cove-oc4ids) we can either disable it like in cove-oc4ids, or we can only calculate it when someone (infrequently) clicks the link to download the Excel file.
jpmckinney commented
zgrep /flattened /var/log/apache2/other_vhosts_access.log*.gz | grep -v 'python-requests/2.24.0' | cut -d' ' -f 5-12 | sort
Shows 32 non-test requests within the last 14 days. So, this should be on-demand, rather than disabled.
jpmckinney commented
cove-ocds already does it on demand (flatten=request.POST.get("flatten")
argument to convert_json
), whereas cove-oc4ids did it always (flatten=True
). So, this issue has no action.