1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
|
/**
* @file _httpinternal.h
* @brief Implementation constants and magic numbers
*
* $LicenseInfo:firstyear=2012&license=viewerlgpl$
* Second Life Viewer Source Code
* Copyright (C) 2012-2013, Linden Research, Inc.
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation;
* version 2.1 of the License only.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*
* Linden Research, Inc., 945 Battery Street, San Francisco, CA 94111 USA
* $/LicenseInfo$
*/
#ifndef _LLCORE_HTTP_INTERNAL_H_
#define _LLCORE_HTTP_INTERNAL_H_
// If you find this included in a public interface header,
// something wrong is probably happening.
// --------------------------------------------------------------------
// General library to-do list
//
// - Implement policy classes. Structure is mostly there just didn't
// need it for the first consumer. [Classes are there. More
// advanced features, like borrowing, aren't there yet.]
// - Consider Removing 'priority' from the request interface. Its use
// in an always active class can lead to starvation of low-priority
// requests. Requires coodination of priority values across all
// components that share a class. Changing priority across threads
// is slightly expensive (relative to gain) and hasn't been completely
// implemented. And the major user of priority, texture fetches,
// may not really need it.
// - Set/get for global policy and policy classes is clumsy. Rework
// it heading in a direction that allows for more dynamic behavior.
// [Mostly fixed]
// - Move HttpOpRequest::prepareRequest() to HttpLibcurl for the
// pedantic.
// - Update downloader and other long-duration services are going to
// need a progress notification. Initial idea is to introduce a
// 'repeating request' which can piggyback on another request and
// persist until canceled or carrier completes. Current queue
// structures allow an HttpOperation object to be enqueued
// repeatedly, so...
// - Investigate making c-ares' re-implementation of a resolver library
// more resilient or more intelligent on Mac. Part of the DNS failure
// lies in here. The mechanism also looks a little less dynamic
// than needed in an environments where networking is changing.
// - Global optimizations: 'borrowing' connections from other classes,
// HTTP pipelining.
// - Dynamic/control system stuff: detect problems and self-adjust.
// This won't help in the face of the router problems we've looked
// at, however. Detect starvation due to UDP activity and provide
// feedback to it.
// - Change the transfer timeout scheme. We're less interested in
// absolute time, in most cases, than in continuous progress.
// - Many of the policy class settings are currently applied to the
// entire class. Some, like connection limits, would be better
// applied to each destination target making multiple targets
// independent.
//
// Integration to-do list
// - LLTextureFetch still needs a major refactor. The use of
// LLQueuedThread makes it hard to inspect workers and do the
// resource waiting we're now doing. Rebuild along simpler lines
// some of which are suggested in new commentary at the top of
// the main source file.
// - Expand areas of usage eventually leading to the removal of LLCurl.
// Rough order of expansion:
// . Avatar names
// . Group membership lists
// . Caps access in general
// . 'The rest'
// - Adapt texture cache, image decode and other image consumers to
// the BufferArray model to reduce data copying. Alternatively,
// adapt this library to something else.
//
// --------------------------------------------------------------------
// If '1', internal ready queues will not order ready
// requests by priority, instead it's first-come-first-served.
// Reprioritization requests have the side-effect of then
// putting the modified request at the back of the ready queue.
#define LLCORE_HTTP_READY_QUEUE_IGNORES_PRIORITY 1
namespace LLCore
{
// Maxium number of policy classes that can be defined.
// *TODO: Currently limited to the default class + 1, extend.
// (TSN: should this be more dynamically sized. Is there a reason to hard limit the number of policies?)
const int HTTP_POLICY_CLASS_LIMIT = 32;
// Debug/informational tracing. Used both
// as a global option and in per-request traces.
const int HTTP_TRACE_OFF = 0;
const int HTTP_TRACE_LOW = 1;
const int HTTP_TRACE_CURL_HEADERS = 2;
const int HTTP_TRACE_CURL_BODIES = 3;
const int HTTP_TRACE_MIN = HTTP_TRACE_OFF;
const int HTTP_TRACE_MAX = HTTP_TRACE_CURL_BODIES;
// Request retry limits
//
// At a minimum, retries need to extend past any throttling
// window we're expecting from central services. In the case
// of Linden services running through the caps routers, there's
// a five-second or so window for throttling with some spillover.
// We want to span a few windows to allow transport to slow
// after onset of the throttles and then recover without a final
// failure. Other systems may need other constants.
const int HTTP_RETRY_COUNT_DEFAULT = 5;
const int HTTP_RETRY_COUNT_MIN = 0;
const int HTTP_RETRY_COUNT_MAX = 100;
const HttpTime HTTP_RETRY_BACKOFF_MIN_DEFAULT = 1E6L; // 1 sec
const HttpTime HTTP_RETRY_BACKOFF_MAX_DEFAULT = 5E6L; // 5 sec
const HttpTime HTTP_RETRY_BACKOFF_MAX = 20E6L; // 20 sec
const int HTTP_REDIRECTS_DEFAULT = 10;
// Timeout value used for both connect and protocol exchange.
// Retries and time-on-queue are not included and aren't
// accounted for.
const long HTTP_REQUEST_TIMEOUT_DEFAULT = 30L;
const long HTTP_REQUEST_XFER_TIMEOUT_DEFAULT = 0L;
const long HTTP_REQUEST_TIMEOUT_MIN = 0L;
const long HTTP_REQUEST_TIMEOUT_MAX = 3600L;
// Limits on connection counts
const int HTTP_CONNECTION_LIMIT_DEFAULT = 8;
const int HTTP_CONNECTION_LIMIT_MIN = 1;
const int HTTP_CONNECTION_LIMIT_MAX = 256;
// Pipelining limits
const long HTTP_PIPELINING_DEFAULT = 0L;
const long HTTP_PIPELINING_MAX = 20L;
// Miscellaneous defaults
const bool HTTP_USE_RETRY_AFTER_DEFAULT = true;
const long HTTP_THROTTLE_RATE_DEFAULT = 0L;
// Tuning parameters
// Time worker thread sleeps after a pass through the
// request, ready and active queues.
const int HTTP_SERVICE_LOOP_SLEEP_NORMAL_MS = 2;
// Block allocation size (a tuning parameter) is found
// in bufferarray.h.
} // end namespace LLCore
#endif // _LLCORE_HTTP_INTERNAL_H_
|
