blob: 6952fc557ced907917d318bdfa401dc909775af5 [file] [log] [blame]
// Copyright (c) 2012 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#ifndef NET_BASE_LOAD_TIMING_INFO_H_
#define NET_BASE_LOAD_TIMING_INFO_H_
#include <stdint.h>
#include "base/time/time.h"
#include "net/base/net_export.h"
namespace net {
// Structure containing timing information for a request.
// It addresses the needs of
// http://groups.google.com/group/http-archive-specification/web/har-1-1-spec,
// http://dev.w3.org/2006/webapi/WebTiming/, and
// http://www.w3.org/TR/resource-timing/.
//
// All events that do not apply to a request have null times. For non-HTTP
// requests, all times other than the request_start times are null.
//
// Requests with connection errors generally only have request start times as
// well, since they never received an established socket.
//
// The general order for events is:
// request_start
// service_worker_start_time
// proxy_start
// proxy_end
// dns_start
// dns_end
// connect_start
// ssl_start
// ssl_end
// connect_end
// send_start
// send_end
// service_worker_ready_time
// service_worker_fetch_start
// service_worker_respond_with_settled
// first_early_hints_time
// receive_headers_start
// receive_headers_end
//
// Times represent when a request starts/stops blocking on an event(*), not the
// time the events actually occurred. In particular, in the case of preconnects
// and socket reuse, no time may be spent blocking on establishing a connection.
// In the case of SPDY, PAC scripts are only run once for each shared session,
// so no time may be spent blocking on them.
//
// (*) Note 1: push_start and push_end are the exception to this, as they
// represent the operation which is asynchronous to normal request flow and
// hence are provided as absolute values and not converted to "blocking" time.
//
// (*) Note 2: Internally to the network stack, times are those of actual event
// occurrence. URLRequest converts them to time which the network stack was
// blocked on each state, as per resource timing specs.
//
// DNS and SSL times are both times for the host, not the proxy, so DNS times
// when using proxies are null, and only requests to HTTPS hosts (Not proxies)
// have SSL times.
struct NET_EXPORT LoadTimingInfo {
// Contains the LoadTimingInfo events related to establishing a connection.
// These are all set by ConnectJobs.
struct NET_EXPORT_PRIVATE ConnectTiming {
ConnectTiming();
~ConnectTiming();
// The time spent looking up the host's DNS address. Null for requests that
// used proxies to look up the DNS address. Also null for SOCKS4 proxies,
// since the DNS address is only looked up after the connection is
// established, which results in unexpected event ordering.
// TODO(mmenke): The SOCKS4 event ordering could be refactored to allow
// these times to be non-null.
// Corresponds to |domainLookupStart| and |domainLookupEnd| in
// ResourceTiming (http://www.w3.org/TR/resource-timing/) for Web-surfacing
// requests.
base::TimeTicks dns_start;
base::TimeTicks dns_end;
// The time spent establishing the connection. Connect time includes proxy
// connect times (though not proxy_resolve or DNS lookup times), time spent
// waiting in certain queues, TCP, and SSL time.
// TODO(mmenke): For proxies, this includes time spent blocking on higher
// level socket pools. Fix this.
// TODO(mmenke): Retried connections to the same server should apparently
// be included in this time. Consider supporting that.
// Since the network stack has multiple notions of a "retry",
// handled at different levels, this may not be worth
// worrying about - backup jobs, reused socket failure,
// multiple round authentication.
// Corresponds to |connectStart| and |connectEnd| in ResourceTiming
// (http://www.w3.org/TR/resource-timing/) for Web-surfacing requests.
base::TimeTicks connect_start;
base::TimeTicks connect_end;
// The time when the SSL handshake started / completed. For non-HTTPS
// requests these are null. These times are only for the SSL connection to
// the final destination server, not an SSL/SPDY proxy.
// |ssl_start| corresponds to |secureConnectionStart| in ResourceTiming
// (http://www.w3.org/TR/resource-timing/) for Web-surfacing requests.
base::TimeTicks ssl_start;
base::TimeTicks ssl_end;
};
LoadTimingInfo();
LoadTimingInfo(const LoadTimingInfo& other);
~LoadTimingInfo();
// True if the socket was reused. When true, DNS, connect, and SSL times
// will all be null. When false, those times may be null, too, for non-HTTP
// requests, or when they don't apply to a request.
//
// For requests that are sent again after an AUTH challenge, this will be true
// if the original socket is reused, and false if a new socket is used.
// Responding to a proxy AUTH challenge is never considered to be reusing a
// socket, since a connection to the host wasn't established when the
// challenge was received.
bool socket_reused;
// Unique socket ID, can be used to identify requests served by the same
// socket. For connections tunnelled over SPDY proxies, this is the ID of
// the virtual connection (The SpdyProxyClientSocket), not the ID of the
// actual socket. HTTP requests handled by the SPDY proxy itself all use the
// actual socket's ID.
//
// 0 when there is no socket associated with the request, or it's not an HTTP
// request.
uint32_t socket_log_id;
// Start time as a base::Time, so times can be coverted into actual times.
// Other times are recorded as TimeTicks so they are not affected by clock
// changes.
base::Time request_start_time;
// Corresponds to |fetchStart| in ResourceTiming
// (http://www.w3.org/TR/resource-timing/) for Web-surfacing requests.
base::TimeTicks request_start;
// The time immediately before starting ServiceWorker. If the response is not
// provided by the ServiceWorker, kept empty.
// Corresponds to |workerStart| in
// ResourceTiming (http://www.w3.org/TR/resource-timing/) for Web-surfacing
base::TimeTicks service_worker_start_time;
// The time immediately before dispatching fetch event in ServiceWorker.
// If the response is not provided by the ServiceWorker, kept empty.
// This value will be used for |fetchStart| (or |redirectStart|) in
// ResourceTiming (http://www.w3.org/TR/resource-timing/) for Web-surfacing
// if this is greater than |request_start|.
base::TimeTicks service_worker_ready_time;
// The time when serviceworker fetch event was popped off the event queue
// and fetch event handler started running.
// If the response is not provided by the ServiceWorker, kept empty.
base::TimeTicks service_worker_fetch_start;
// The time when serviceworker's fetch event's respondWith promise was
// settled. If the response is not provided by the ServiceWorker, kept empty.
base::TimeTicks service_worker_respond_with_settled;
// The time spent determining which proxy to use. Null when there is no PAC.
base::TimeTicks proxy_resolve_start;
base::TimeTicks proxy_resolve_end;
ConnectTiming connect_timing;
// The time that sending HTTP request started / ended.
// |send_start| corresponds to |requestStart| in ResourceTiming
// (http://www.w3.org/TR/resource-timing/) for Web-surfacing requests.
base::TimeTicks send_start;
base::TimeTicks send_end;
// The time at which the first / last byte of the HTTP headers were received.
// |receive_headers_start| corresponds to |responseStart| in ResourceTiming
// (http://www.w3.org/TR/resource-timing/) for Web-surfacing requests.
base::TimeTicks receive_headers_start;
base::TimeTicks receive_headers_end;
// The time that the first 103 Early Hints response is received.
base::TimeTicks first_early_hints_time;
// In case the resource was proactively pushed by the server, these are
// the times that push started and ended. Note that push_end will be null
// if the request is still being transmitted, i.e. the underlying h2 stream
// is not closed by the server.
base::TimeTicks push_start;
base::TimeTicks push_end;
};
} // namespace net
#endif // NET_BASE_LOAD_TIMING_INFO_H_