Fifteen minutes of testing before payment saves a month of working with a set that was never going to fit.
An address-reporting service tells you the address is alive and nothing else. It says nothing about behaviour under load.
Run a trial of a thousand records against the platform you are buying the addresses for, at the pace you intend to use.
The failure rate towards the end of the run: if it climbs, the set is one range. The spread of response times: one slow address holds up everything.
And behaviour at a hundred threads: some sets hold ten and fall apart at a hundred.
Group the addresses by the first three numbers. The number of groups is the spread the vendor described.
Check the whole list, all the way down: sometimes the first half matches the description and the second arrived from a neighbouring country.
How many subnets are in the set, how quickly problem addresses are replaced, and whether there is a trial. Three answers and the picture is clear.
for a in $(cat pool.txt); do
curl -s -o /dev/null -w "$a %{http_code} %{time_total}\n" -x $a https://example.org/
done | sort -k3 -n | tail -5Сервис не оплатил переходы с площадки, поэтому кнопка перехода к нему закрыта. Место в рейтинге от оплаты не зависит и считается по общей формуле.
Карточка сервиса с характеристиками остаётся открытой, её видно на странице сервиса.