bigQuery HTTP Archive - requests

third-party-calls.sql

/* WIP - not working yet */

SELECT
  NTH(10, num_requests) thenth,
	NTH(25, num_requests) twenty_fifth,
	NTH(50, num_requests) median,
	NTH(75, num_requests) seventy_fifth,
	NTH(90, num_requests) ninetieth
FROM(
	SELECT COUNT(*) num_requests
	FROM [httparchive:runs.latest_requests] as req JOIN (
		SELECT DOMAIN(url) self
		FROM [httparchive:runs.latest_pages]
	) as pages ON pages.self = req.req_host
	WHERE DOMAIN(req.url) != pages.self
	GROUP BY pages.self
	ORDER BY num_requests asc
)

total-domains.sql

SELECT
  NTH(25, quantiles(numDomains,100)) twenty_fifth,
  NTH(50, quantiles(numDomains,100)) median,
  NTH(75, quantiles(numDomains,100)) seventy_fifth,
  NTH(90, quantiles(numDomains,100)) ninetieth
FROM [httparchive:runs.latest_pages]

total-requests.sql

SELECT
  NTH(25, quantiles(reqTotal,100)) twenty_fifth,
  NTH(50, quantiles(reqTotal,100)) median,
  NTH(75, quantiles(reqTotal,100)) seventy_fifth,
  NTH(90, quantiles(reqTotal,100)) ninetieth
FROM [httparchive:runs.latest_pages]

https://bigquery.cloud.google.com/

Full description:
https://www.igvita.com/2013/06/20/http-archive-bigquery-web-performance-answers
Some inspiration:
https://gist.github.com/igrigorik/5801492