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
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
|
/*
* Copyright (c) 2017-2021 Cisco and/or its affiliates.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at:
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include <glog/logging.h>
#include <protocols/rtc/rtc_consts.h>
#include <protocols/rtc/rtc_ldr.h>
#include <algorithm>
#include <unordered_set>
namespace transport {
namespace protocol {
namespace rtc {
RTCLossDetectionAndRecovery::RTCLossDetectionAndRecovery(
Indexer *indexer, SendRtxCallback &&callback, asio::io_service &io_service)
: rtx_on_(false),
fec_on_(false),
next_rtx_timer_(MAX_TIMER_RTX),
last_event_(0),
sentinel_timer_interval_(MAX_TIMER_RTX),
indexer_(indexer),
send_rtx_callback_(std::move(callback)) {
timer_ = std::make_unique<asio::steady_timer>(io_service);
sentinel_timer_ = std::make_unique<asio::steady_timer>(io_service);
}
RTCLossDetectionAndRecovery::~RTCLossDetectionAndRecovery() {}
void RTCLossDetectionAndRecovery::turnOnRTX() {
rtx_on_ = true;
scheduleSentinelTimer(state_->getRTT() * CATCH_UP_RTT_INCREMENT);
}
void RTCLossDetectionAndRecovery::turnOffRTX() {
rtx_on_ = false;
clear();
}
uint32_t RTCLossDetectionAndRecovery::computeFecPacketsToAsk(bool in_sync) {
uint32_t current_fec = indexer_->getNFec();
double current_loss_rate = state_->getLossRate();
double last_loss_rate = state_->getLastRoundLossRate();
// when in sync ask for fec only if there are losses for 2 rounds
if (in_sync && current_fec == 0 &&
(current_loss_rate == 0 || last_loss_rate == 0))
return 0;
double loss_rate = state_->getMaxLossRate() * 1.5;
if (!in_sync && loss_rate == 0) loss_rate = 0.05;
if (loss_rate > 0.5) loss_rate = 0.5;
double exp_losses = (double)k_ * loss_rate;
uint32_t fec_to_ask = ceil(exp_losses / (1 - loss_rate));
if (fec_to_ask > (n_ - k_)) fec_to_ask = n_ - k_;
return fec_to_ask;
}
void RTCLossDetectionAndRecovery::onNewRound(bool in_sync) {
uint64_t rtt = state_->getRTT();
if (!fec_on_ && rtt >= 100) {
// turn on fec, here we may have no info so ask for all packets
fec_on_ = true;
turnOffRTX();
indexer_->setNFec(computeFecPacketsToAsk(in_sync));
return;
}
if (fec_on_ && rtt > 80) {
// keep using fec, maybe update it
indexer_->setNFec(computeFecPacketsToAsk(in_sync));
return;
}
if ((fec_on_ && rtt <= 80) || (!rtx_on_ && rtt <= 100)) {
// turn on rtx
fec_on_ = false;
indexer_->setNFec(0);
turnOnRTX();
return;
}
}
void RTCLossDetectionAndRecovery::onTimeout(uint32_t seq) {
// always add timeouts to the RTX list to avoid to send the same packet as if
// it was not a rtx
addToRetransmissions(seq, seq + 1);
last_event_ = getNow();
}
void RTCLossDetectionAndRecovery::onPacketRecoveredFec(uint32_t seq) {
// if an RTX is scheduled for a packet recovered using FEC delete it
deleteRtx(seq);
recover_with_fec_.erase(seq);
}
void RTCLossDetectionAndRecovery::onDataPacketReceived(
const core::ContentObject &content_object) {
last_event_ = getNow();
uint32_t seq = content_object.getName().getSuffix();
if (deleteRtx(seq)) {
state_->onPacketRecoveredRtx(seq);
} else {
DLOG_IF(INFO, VLOG_IS_ON(3))
<< "received data. add from "
<< state_->getHighestSeqReceivedInOrder() + 1 << " to " << seq;
addToRetransmissions(state_->getHighestSeqReceivedInOrder() + 1, seq);
}
}
void RTCLossDetectionAndRecovery::onNackPacketReceived(
const core::ContentObject &nack) {
last_event_ = getNow();
uint32_t seq = nack.getName().getSuffix();
struct nack_packet_t *nack_pkt =
(struct nack_packet_t *)nack.getPayload()->data();
uint32_t production_seq = nack_pkt->getProductionSegement();
if (production_seq > seq) {
// this is a past nack, all data before productionSeq are lost. if
// productionSeq > state_->getHighestSeqReceivedInOrder() is impossible to
// recover any packet. If this is not the case we can try to recover the
// packets between state_->getHighestSeqReceivedInOrder() and productionSeq.
// e.g.: the client receives packets 8 10 11 9 where 9 is a nack with
// productionSeq = 14. 9 is lost but we can try to recover packets 12 13 and
// 14 that are not arrived yet
deleteRtx(seq);
DLOG_IF(INFO, VLOG_IS_ON(3)) << "received past nack. add from "
<< state_->getHighestSeqReceivedInOrder() + 1
<< " to " << production_seq;
addToRetransmissions(state_->getHighestSeqReceivedInOrder() + 1,
production_seq);
} else {
// future nack. here there should be a gap between the last data received
// and this packet and is it possible to recover the packets between the
// last received data and the production seq. we should not use the seq
// number of the nack since we know that is too early to ask for this seq
// number
// e.g.: // e.g.: the client receives packets 10 11 12 20 where 20 is a nack
// with productionSeq = 18. this says that all the packets between 12 and 18
// may got lost and we should ask them
deleteRtx(seq);
DLOG_IF(INFO, VLOG_IS_ON(3)) << "received futrue nack. add from "
<< state_->getHighestSeqReceivedInOrder() + 1
<< " to " << production_seq;
addToRetransmissions(state_->getHighestSeqReceivedInOrder() + 1,
production_seq);
}
}
void RTCLossDetectionAndRecovery::onProbePacketReceived(
const core::ContentObject &probe) {
// we don't log the reception of a probe packet for the sentinel timer because
// probes are not taken into account into the sync window. we use them as
// future nacks to detect possible packets lost
struct nack_packet_t *probe_pkt =
(struct nack_packet_t *)probe.getPayload()->data();
uint32_t production_seq = probe_pkt->getProductionSegement();
DLOG_IF(INFO, VLOG_IS_ON(3))
<< "received probe. add from "
<< state_->getHighestSeqReceivedInOrder() + 1 << " to " << production_seq;
addToRetransmissions(state_->getHighestSeqReceivedInOrder() + 1,
production_seq);
}
void RTCLossDetectionAndRecovery::clear() {
rtx_state_.clear();
rtx_timers_.clear();
sentinel_timer_->cancel();
if (next_rtx_timer_ != MAX_TIMER_RTX) {
next_rtx_timer_ = MAX_TIMER_RTX;
timer_->cancel();
}
}
void RTCLossDetectionAndRecovery::addToRetransmissions(uint32_t start,
uint32_t stop) {
// skip nacked packets
if (start <= state_->getLastSeqNacked()) {
start = state_->getLastSeqNacked() + 1;
}
// skip received or lost packets
if (start <= state_->getHighestSeqReceivedInOrder()) {
start = state_->getHighestSeqReceivedInOrder() + 1;
}
for (uint32_t seq = start; seq < stop; seq++) {
if (state_->isReceivedOrLost(seq) == PacketState::UNKNOWN) {
if (rtx_on_) {
if (!indexer_->isFec(seq)) {
// handle it with rtx
if (!isRtx(seq)) {
state_->onLossDetected(seq);
rtxState state;
state.first_send_ = state_->getInterestSentTime(seq);
if (state.first_send_ == 0) // this interest was never sent before
state.first_send_ = getNow();
state.next_send_ = computeNextSend(seq, true);
state.rtx_count_ = 0;
DLOG_IF(INFO, VLOG_IS_ON(4))
<< "Add " << seq << " to retransmissions. next rtx is %lu "
<< state.next_send_ - getNow();
rtx_state_.insert(std::pair<uint32_t, rtxState>(seq, state));
rtx_timers_.insert(
std::pair<uint64_t, uint32_t>(state.next_send_, seq));
}
} else {
// is fec, do not send it
auto it = recover_with_fec_.find(seq);
if (it == recover_with_fec_.end()) {
state_->onLossDetected(seq);
recover_with_fec_.insert(seq);
}
}
} else {
// keep track of losses but recover with FEC
auto it = recover_with_fec_.find(seq);
if (it == recover_with_fec_.end()) {
state_->onLossDetected(seq);
recover_with_fec_.insert(seq);
}
}
}
}
scheduleNextRtx();
}
uint64_t RTCLossDetectionAndRecovery::computeNextSend(uint32_t seq,
bool new_rtx) {
uint64_t now = getNow();
if (new_rtx) {
// for the new rtx we wait one estimated IAT after the loss detection. this
// is bacause, assuming that packets arrive with a constant IAT, we should
// get a new packet every IAT
double prod_rate = state_->getProducerRate();
uint32_t estimated_iat = SENTINEL_TIMER_INTERVAL;
uint32_t jitter = 0;
if (prod_rate != 0) {
double packet_size = state_->getAveragePacketSize();
estimated_iat = ceil(1000.0 / (prod_rate / packet_size));
jitter = ceil(state_->getJitter());
}
uint32_t wait = estimated_iat + jitter;
DLOG_IF(INFO, VLOG_IS_ON(3))
<< "first rtx for " << seq << " in " << wait
<< " ms, rtt = " << state_->getRTT() << " ait = " << estimated_iat
<< " jttr = " << jitter;
return now + wait;
} else {
// wait one RTT
// however if the IAT is larger than the RTT, wait one IAT
uint32_t wait = SENTINEL_TIMER_INTERVAL;
double prod_rate = state_->getProducerRate();
if (prod_rate == 0) {
return now + SENTINEL_TIMER_INTERVAL;
}
double packet_size = state_->getAveragePacketSize();
uint32_t estimated_iat = ceil(1000.0 / (prod_rate / packet_size));
uint64_t rtt = state_->getRTT();
if (rtt == 0) rtt = SENTINEL_TIMER_INTERVAL;
wait = rtt;
if (estimated_iat > rtt) wait = estimated_iat;
uint32_t jitter = ceil(state_->getJitter());
wait += jitter;
// it may happen that the channel is congested and we have some additional
// queuing delay to take into account
uint32_t queue = ceil(state_->getQueuing());
wait += queue;
DLOG_IF(INFO, VLOG_IS_ON(3))
<< "next rtx for " << seq << " in " << wait
<< " ms, rtt = " << state_->getRTT() << " ait = " << estimated_iat
<< " jttr = " << jitter << " queue = " << queue;
return now + wait;
}
}
void RTCLossDetectionAndRecovery::retransmit() {
if (rtx_timers_.size() == 0) return;
uint64_t now = getNow();
auto it = rtx_timers_.begin();
std::unordered_set<uint32_t> lost_pkt;
uint32_t sent_counter = 0;
while (it != rtx_timers_.end() && it->first <= now &&
sent_counter < MAX_RTX_IN_BATCH) {
uint32_t seq = it->second;
auto rtx_it =
rtx_state_.find(seq); // this should always return a valid iter
if (rtx_it->second.rtx_count_ >= RTC_MAX_RTX ||
(now - rtx_it->second.first_send_) >= RTC_MAX_AGE ||
seq < state_->getLastSeqNacked()) {
// max rtx reached or packet too old or packet nacked, this packet is lost
DLOG_IF(INFO, VLOG_IS_ON(3))
<< "packet " << seq << " lost because 1) max rtx: "
<< (rtx_it->second.rtx_count_ >= RTC_MAX_RTX) << " 2) max age: "
<< ((now - rtx_it->second.first_send_) >= RTC_MAX_AGE)
<< " 3) nacked: " << (seq < state_->getLastSeqNacked());
lost_pkt.insert(seq);
it++;
} else {
// resend the packet
state_->onRetransmission(seq);
double prod_rate = state_->getProducerRate();
if (prod_rate != 0) rtx_it->second.rtx_count_++;
rtx_it->second.next_send_ = computeNextSend(seq, false);
it = rtx_timers_.erase(it);
rtx_timers_.insert(
std::pair<uint64_t, uint32_t>(rtx_it->second.next_send_, seq));
DLOG_IF(INFO, VLOG_IS_ON(3))
<< "send rtx for sequence " << seq << ", next send in "
<< (rtx_it->second.next_send_ - now);
send_rtx_callback_(seq);
sent_counter++;
}
}
// remove packets if needed
for (auto lost_it = lost_pkt.begin(); lost_it != lost_pkt.end(); lost_it++) {
uint32_t seq = *lost_it;
state_->onPacketLost(seq);
deleteRtx(seq);
}
}
void RTCLossDetectionAndRecovery::scheduleNextRtx() {
if (rtx_timers_.size() == 0) {
// all the rtx were removed, reset timer
next_rtx_timer_ = MAX_TIMER_RTX;
return;
}
// check if timer is alreay set
if (next_rtx_timer_ != MAX_TIMER_RTX) {
// a new check for rtx is already scheduled
if (next_rtx_timer_ > rtx_timers_.begin()->first) {
// we need to re-schedule it
timer_->cancel();
} else {
// wait for the next timer
return;
}
}
// set a new timer
next_rtx_timer_ = rtx_timers_.begin()->first;
uint64_t now = std::chrono::duration_cast<std::chrono::milliseconds>(
std::chrono::steady_clock::now().time_since_epoch())
.count();
uint64_t wait = 1;
if (next_rtx_timer_ != MAX_TIMER_RTX && next_rtx_timer_ > now)
wait = next_rtx_timer_ - now;
std::weak_ptr<RTCLossDetectionAndRecovery> self(shared_from_this());
timer_->expires_from_now(std::chrono::milliseconds(wait));
timer_->async_wait([self](std::error_code ec) {
if (ec) return;
if (auto s = self.lock()) {
s->retransmit();
s->next_rtx_timer_ = MAX_TIMER_RTX;
s->scheduleNextRtx();
}
});
}
bool RTCLossDetectionAndRecovery::deleteRtx(uint32_t seq) {
auto it_rtx = rtx_state_.find(seq);
if (it_rtx == rtx_state_.end()) return false; // rtx not found
uint64_t ts = it_rtx->second.next_send_;
auto it_timers = rtx_timers_.find(ts);
while (it_timers != rtx_timers_.end() && it_timers->first == ts) {
if (it_timers->second == seq) {
rtx_timers_.erase(it_timers);
break;
}
it_timers++;
}
bool lost = it_rtx->second.rtx_count_ > 0;
rtx_state_.erase(it_rtx);
return lost;
}
void RTCLossDetectionAndRecovery::scheduleSentinelTimer(
uint64_t expires_from_now) {
std::weak_ptr<RTCLossDetectionAndRecovery> self(shared_from_this());
sentinel_timer_->expires_from_now(
std::chrono::milliseconds(expires_from_now));
sentinel_timer_->async_wait([self](std::error_code ec) {
if (ec) return;
if (auto s = self.lock()) {
s->sentinelTimer();
}
});
}
void RTCLossDetectionAndRecovery::sentinelTimer() {
uint64_t now = getNow();
bool expired = false;
bool sent = false;
if ((now - last_event_) >= sentinel_timer_interval_) {
// at least a sentinel_timer_interval_ elapsed since last event
expired = true;
if (TRANSPORT_EXPECT_FALSE(!state_->isProducerActive())) {
// this happens at the beginning (or if the producer stops for some
// reason) we need to keep sending interest 0 until we get an answer
DLOG_IF(INFO, VLOG_IS_ON(3))
<< "sentinel timer: the producer is not active, send packet 0";
state_->onRetransmission(0);
send_rtx_callback_(0);
} else {
DLOG_IF(INFO, VLOG_IS_ON(3)) << "sentinel timer: the producer is active, "
"send the 10 oldest packets";
sent = true;
uint32_t rtx = 0;
auto it = state_->getPendingInterestsMapBegin();
auto end = state_->getPendingInterestsMapEnd();
while (it != end && rtx < MAX_RTX_WITH_SENTINEL) {
uint32_t seq = it->first;
DLOG_IF(INFO, VLOG_IS_ON(3))
<< "sentinel timer, add " << seq << " to the rtx list";
addToRetransmissions(seq, seq + 1);
rtx++;
it++;
}
}
} else {
// sentinel timer did not expire because we registered at least one event
}
uint32_t next_timer;
double prod_rate = state_->getProducerRate();
if (TRANSPORT_EXPECT_FALSE(!state_->isProducerActive()) || prod_rate == 0) {
DLOG_IF(INFO, VLOG_IS_ON(3)) << "next timer in " << SENTINEL_TIMER_INTERVAL;
next_timer = SENTINEL_TIMER_INTERVAL;
} else {
double prod_rate = state_->getProducerRate();
double packet_size = state_->getAveragePacketSize();
uint32_t estimated_iat = ceil(1000.0 / (prod_rate / packet_size));
uint32_t jitter = ceil(state_->getJitter());
// try to reduce the number of timers if the estimated IAT is too small
next_timer = std::max((estimated_iat + jitter) * 20, (uint32_t)1);
DLOG_IF(INFO, VLOG_IS_ON(3))
<< "next sentinel in " << next_timer
<< " ms, rate: " << ((prod_rate * 8.0) / 1000000.0)
<< ", iat: " << estimated_iat << ", jitter: " << jitter;
if (!expired) {
// discount the amout of time that is already passed
uint32_t discount = now - last_event_;
if (next_timer > discount) {
next_timer = next_timer - discount;
} else {
// in this case we trigger the timer in 1 ms
next_timer = 1;
}
DLOG_IF(INFO, VLOG_IS_ON(3)) << "timer after discout: " << next_timer;
} else if (sent) {
// wait at least one producer stats interval + owd to check if the
// production rate is reducing.
uint32_t min_wait = PRODUCER_STATS_INTERVAL + ceil(state_->getQueuing());
next_timer = std::max(next_timer, min_wait);
DLOG_IF(INFO, VLOG_IS_ON(3))
<< "wait for updates from prod, next timer: " << next_timer;
}
}
scheduleSentinelTimer(next_timer);
}
} // namespace rtc
} // namespace protocol
} // namespace transport
|