-
Notifications
You must be signed in to change notification settings - Fork 4k
/
Copy pathrpl_io_monitor.cc
1152 lines (977 loc) · 39.9 KB
/
rpl_io_monitor.cc
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
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/* Copyright (c) 2020, 2024, Oracle and/or its affiliates.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License, version 2.0,
as published by the Free Software Foundation.
This program is designed to work with certain software (including
but not limited to OpenSSL) that is licensed under separate terms,
as designated in a particular file or component or in included license
documentation. The authors of MySQL hereby grant you an additional
permission to link the program and your derivative works with the
separately licensed software that they have either included with
the program or referenced in the documentation.
This program 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 General Public License, version 2.0, for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */
#include "sql/rpl_io_monitor.h"
#include <mysql/components/my_service.h>
#include <mysql/components/services/group_replication_status_service.h>
#include "mysql/components/services/log_builtins.h"
#include "sql-common/json_dom.h"
#include "sql/changestreams/apply/replication_thread_status.h"
#include "sql/mysqld.h"
#include "sql/mysqld_thd_manager.h" // Global_THD_manager
#include "sql/protocol_classic.h"
#include "sql/rpl_async_conn_failover.h" // reset_pos
#include "sql/rpl_async_conn_failover_configuration_propagation.h"
#include "sql/rpl_group_replication.h"
#include "sql/rpl_msr.h" /* Multisource replication */
#include "sql/rpl_replica.h"
#include "sql/rpl_sys_key_access.h"
#include "sql/rpl_sys_table_access.h"
#include "sql/sql_class.h" // THD
#include "sql/udf_service_util.h"
#include "my_dbug.h"
#include "my_systime.h"
#include <string>
/**
Restart the IO thread of the given channel.
@param[in] thd The running thread.
@param[in] channel_name the channel IO thread to restart.
@param[in] force_sender_with_highest_weight When true, sender with highest
weight is chosen, otherwise the next sender from the current one is chosen.
@return true if IO thread was restarted, false otherwise.
*/
static bool restart_io_thread(THD *thd, const std::string &channel_name,
bool force_sender_with_highest_weight);
/*
The SQL_QUERIES array contains three queries. The enum_sql_query_tag index/tag
is used to get each query. There are for following purpose:
1. CONFIG_MODE_QUORUM_MONITOR:
Its used by Monitor IO thread to determine if given source has Group
Replication enabled and if enabled whether member is in ONLINE or
RECOVERING state and has QUORUM.
2. CONFIG_MODE_QUORUM_IO:
Its used by IO thread to determine if given source has Group Replication
enabled and if enabled whether member is in ONLINE state and has QUORUM.
3. GR_MEMBER_ALL_DETAILS:
Its used by Monitor IO thread to get following member details:
group_name, host, port, member state and member role.
4. GR_MEMBER_ALL_DETAILS_FETCH_FOR_57:
Its used by Monitor IO thread for mysql-5.7 servers to get following
member details:
group_name, host, port, member state and member role.
In mysql-5.7 performance_schema.replication_group_members do not have
member role column but its fetched from group_replication_primary_member
status variable, when group is on single-primary mode.
5. QUERY_SERVER_SELECT_ONE:
Its used by Monitor IO thread to check single-server is in working state.
It establishes connection with single server and executes this
query to confirm that connection to SOURCE is working.
*/
static const char *SQL_QUERIES[] = {
"SELECT * FROM ( "
" SELECT CASE "
" WHEN ((SELECT count(*) from information_schema.plugins WHERE "
" PLUGIN_NAME LIKE 'group_replication') <> 1) "
" THEN (SELECT 2) "
" WHEN ((SELECT IF(((MEMBER_STATE='ONLINE') OR "
" (MEMBER_STATE='RECOVERING')) AND "
" ((SELECT COUNT(*) FROM "
" performance_schema.replication_group_members "
" WHERE MEMBER_STATE != 'ONLINE' AND MEMBER_STATE != "
"'RECOVERING') "
" >= ((SELECT COUNT(*) FROM "
" performance_schema.replication_group_members)/2)=0),1,0) "
" FROM performance_schema.replication_group_members "
" WHERE member_id=@@global.server_uuid) = 1) "
" THEN (SELECT 1) "
" ELSE (SELECT 2) "
" END AS QUORUM "
") Q ",
"SELECT * FROM ( "
" SELECT CASE "
" WHEN ((SELECT count(*) from information_schema.plugins WHERE "
" PLUGIN_NAME LIKE 'group_replication') <> 1) "
" THEN (SELECT 2) "
" WHEN ((SELECT IF(MEMBER_STATE='ONLINE' AND "
" ((SELECT COUNT(*) FROM "
" performance_schema.replication_group_members "
" WHERE MEMBER_STATE != 'ONLINE' AND MEMBER_STATE != "
"'RECOVERING') "
" >= ((SELECT COUNT(*) FROM "
" performance_schema.replication_group_members)/2)=0),1,0) "
" FROM performance_schema.replication_group_members "
" WHERE member_id=@@global.server_uuid) = 1) "
" THEN (SELECT 1) "
" ELSE (SELECT 2) "
" END AS QUORUM "
") Q ",
"SELECT @@global.group_replication_group_name, PRGM.MEMBER_HOST, "
" PRGM.MEMBER_PORT, PRGM.MEMBER_STATE, PRGM.MEMBER_ROLE "
"FROM performance_schema.replication_group_members PRGM",
"SELECT @@global.group_replication_group_name, PRGM.MEMBER_HOST, "
" PRGM.MEMBER_PORT, PRGM.MEMBER_STATE, "
" (SELECT IF(GR_SINGLE_PRIMARY_MODE.VARIABLE_VALUE = 'OFF', "
" 'PRIMARY', "
" IF(PRGM.MEMBER_ID = GR_PRIMARY_MEMBER.VARIABLE_VALUE, "
" 'PRIMARY', 'SECONDARY')) "
" FROM (SELECT VARIABLE_VALUE FROM performance_schema.global_status "
" WHERE VARIABLE_NAME = 'group_replication_primary_member') "
" GR_PRIMARY_MEMBER,"
" (SELECT VARIABLE_VALUE FROM "
" performance_schema.global_variables "
" WHERE "
" VARIABLE_NAME='group_replication_single_primary_mode') "
" GR_SINGLE_PRIMARY_MODE "
" ) MEMBER_ROLE "
"FROM performance_schema.replication_group_members PRGM",
"SELECT 1"};
MYSQL_RES_TUPLE execute_query(const Mysql_connection *conn,
enum_sql_query_tag qtag) {
int tag = static_cast<int>(qtag);
std::string query = SQL_QUERIES[tag];
return conn->execute_query(query);
}
static void *launch_handler_thread(void *arg) {
Source_IO_monitor *monitor = (Source_IO_monitor *)arg;
monitor->source_monitor_handler();
return nullptr;
}
std::string Source_IO_monitor::get_query(enum_sql_query_tag qtag) {
int tag = static_cast<int>(qtag);
std::string query = SQL_QUERIES[tag];
return query;
}
Source_IO_monitor::Source_IO_monitor() {
#ifdef HAVE_PSI_INTERFACE
mysql_mutex_init(key_monitor_info_run_lock, &m_run_lock, MY_MUTEX_INIT_FAST);
mysql_cond_init(key_monitor_info_run_cond, &m_run_cond);
#else
mysql_mutex_init(nullptr, &m_run_lock, MY_MUTEX_INIT_FAST);
mysql_cond_init(nullptr, &m_run_cond);
#endif
}
Source_IO_monitor::~Source_IO_monitor() {
terminate_monitoring_process();
mysql_mutex_destroy(&m_run_lock);
mysql_cond_destroy(&m_run_cond);
}
bool Source_IO_monitor::is_monitor_killed(THD *thd, Master_info *) {
DBUG_TRACE;
assert(m_monitor_thd == thd);
return m_abort_monitor || connection_events_loop_aborted() || thd->killed;
}
bool Source_IO_monitor::launch_monitoring_process(PSI_thread_key thread_key) {
DBUG_TRACE;
mysql_mutex_lock(&m_run_lock);
// Callers should ensure the process is terminated
assert(!m_monitor_thd_state.is_thread_alive());
if (m_monitor_thd_state.is_thread_alive()) {
mysql_mutex_unlock(&m_run_lock);
return true;
}
if (mysql_thread_create(thread_key, &m_th, &connection_attrib,
launch_handler_thread, (void *)this)) {
my_error(ER_REPLICA_THREAD, MYF(0));
mysql_mutex_unlock(&m_run_lock);
return true;
}
m_monitor_thd_state.set_created();
while (m_monitor_thd_state.is_alive_not_running()) {
DBUG_PRINT("sleep", ("Waiting for the Monitoring process thread to start"));
struct timespec abstime;
set_timespec(&abstime, 1);
mysql_cond_timedwait(&m_run_cond, &m_run_lock, &abstime);
}
mysql_mutex_unlock(&m_run_lock);
return false;
}
void Source_IO_monitor::source_monitor_handler() {
THD *thd{nullptr}; // needs to be first for thread_stack
thd = new THD; // note that constructor of THD uses DBUG_ !
m_monitor_thd = thd;
struct timespec waittime;
DBUG_TRACE;
THD_CHECK_SENTRY(thd);
my_thread_init();
#ifdef HAVE_PSI_THREAD_INTERFACE
// save the instrumentation for IO thread in mi->info_thd
struct PSI_thread *psi = PSI_THREAD_CALL(get_thread)();
thd_set_psi(thd, psi);
#endif
thd->thread_stack = (char *)&thd; // remember where our stack is
if (init_replica_thread(thd, SLAVE_THD_IO)) {
my_error(ER_REPLICA_FATAL_ERROR, MYF(0),
"Failed during Replica IO Monitor thread initialization ");
goto err;
}
thd->security_context()->skip_grants();
Global_THD_manager::get_instance()->add_thd(thd);
mysql_mutex_lock(&m_run_lock);
m_monitor_thd_state.set_running();
m_abort_monitor = false;
set_timespec(&waittime, m_retry_monitor_wait);
mysql_cond_broadcast(&m_run_cond);
mysql_mutex_unlock(&m_run_lock);
while (!is_monitor_killed(thd, nullptr) &&
!is_group_replication_member_secondary()) {
sync_senders_details(thd);
THD_STAGE_INFO(thd, stage_rpl_failover_wait_before_next_fetch);
set_timespec(&waittime, m_retry_monitor_wait);
mysql_mutex_lock(&m_run_lock);
mysql_cond_timedwait(&m_run_cond, &m_run_lock, &waittime);
mysql_mutex_unlock(&m_run_lock);
}
err:
LogErr(INFORMATION_LEVEL, ER_RPL_REPLICA_MONITOR_IO_THREAD_EXITING);
/* At this point the I/O thread will not try to reconnect anymore. */
thd->reset_query();
thd->reset_db(NULL_CSTR);
// destructor will not free it, because net.vio is 0
thd->get_protocol_classic()->end_net();
thd->release_resources();
THD_CHECK_SENTRY(thd);
Global_THD_manager::get_instance()->remove_thd(thd);
delete thd;
mysql_mutex_lock(&m_run_lock);
m_monitor_thd_state.set_terminated();
m_abort_monitor = true;
mysql_cond_broadcast(&m_run_cond);
mysql_mutex_unlock(&m_run_lock);
my_thread_end();
my_thread_exit(nullptr);
return;
}
std::tuple<bool, std::string> Source_IO_monitor::delete_rows(
Rpl_sys_table_access &table_op, TABLE *table,
std::vector<std::string> field_name,
std::tuple<std::string, std::string, uint> conn_detail) {
bool err_val{false};
std::string err_msg{};
Rpl_sys_table_access::for_each_in_tuple(
conn_detail, [&](const auto &n, const auto &x) {
if (table_op.store_field(table->field[n], x)) {
err_msg.assign(table_op.get_field_error_msg(field_name[n]));
err_val = true;
}
});
if (err_val) return std::make_tuple(err_val, err_msg);
Rpl_sys_table_access::handler_delete_row_func(table_op, err_val, err_msg);
return std::make_tuple(err_val, err_msg);
}
std::tuple<bool, std::string> Source_IO_monitor::write_rows(
Rpl_sys_table_access &table_op, TABLE *table,
std::vector<std::string> field_name,
RPL_FAILOVER_SOURCE_TUPLE conn_detail) {
bool err_val{false};
std::string err_msg{};
Rpl_sys_table_access::for_each_in_tuple(
conn_detail, [&](const auto &n, const auto &x) {
if (table_op.store_field(table->field[n], x)) {
err_msg.assign(table_op.get_field_error_msg(field_name[n]));
err_val = true;
}
});
if (err_val) return std::make_tuple(err_val, err_msg);
Rpl_sys_table_access::handler_write_row_func(table_op, err_val, err_msg);
return std::make_tuple(err_val, err_msg);
}
int Source_IO_monitor::connect_senders(THD *thd,
const std::string &channel_name) {
std::vector<SENDER_CONN_MERGE_TUPLE> failover_table_detail_list{};
std::vector<RPL_FAILOVER_SOURCE_TUPLE> source_conn_detail_list{};
bool error{false};
/* highest group failover weight for the current channel. */
uint curr_highest_group_weight{0};
/* weight of single senders for the current channel. */
uint curr_highest_weight_single_sender{0};
/* weight for current connected sender */
uint curr_conn_weight{0};
if (is_monitor_killed(thd, nullptr)) return 1;
/*
1. Get stored source details for channel from
replication_asynchronous_connection_failover table.
*/
std::tie(error, failover_table_detail_list) =
get_senders_details(channel_name);
if (error) {
return 2;
}
if (is_monitor_killed(thd, nullptr)) return 1;
/*
2. Get weight of current connected sender.
*/
{
Rpl_async_conn_failover_table_operations table_op_src(TL_READ);
std::tie(error, source_conn_detail_list) =
table_op_src.read_source_rows_for_channel(channel_name);
std::sort(source_conn_detail_list.begin(), source_conn_detail_list.end(),
[](const RPL_FAILOVER_SOURCE_TUPLE &element1,
const RPL_FAILOVER_SOURCE_TUPLE &element2) -> bool {
return std::get<4>(element1) > std::get<4>(element2);
});
channel_map.rdlock();
Master_info *mi = channel_map.get_mi(channel_name.c_str());
if (nullptr == mi) {
channel_map.unlock();
return 2;
}
const std::string mi_host{mi->host};
const uint mi_port = mi->port;
channel_map.unlock();
for (auto source_conn_detail : source_conn_detail_list) {
uint port{0}, weight{0};
std::string host{""};
std::tie(std::ignore, host, port, std::ignore, weight, std::ignore) =
source_conn_detail;
/* save weight for current connected sender */
if ((host.compare(mi_host) == 0) && (port == mi_port)) {
curr_conn_weight = weight;
break;
}
}
}
/*
3. Connect to one of the source from group gathered in step 1, store their
connection object and get group membership details from the source.
*/
std::set<std::string> managed_name_list{};
for (auto failover_table_detail : failover_table_detail_list) {
uint primary_weight{0}, secondary_weight{0}, port{0}, weight{0};
std::string channel{}, host{}, managed_name{};
std::tie(channel, host, port, std::ignore, weight, managed_name,
primary_weight, secondary_weight) = failover_table_detail;
if (is_monitor_killed(thd, nullptr)) return 1;
/*
3.1. To get group membership details, need to connect to only one member
from the group.
*/
if (managed_name_list.find(managed_name) != managed_name_list.end())
continue;
/* 3.2. Connect to source and store its connection object. */
channel_map.rdlock();
Master_info *mi = channel_map.get_mi(channel_name.c_str());
if (nullptr == mi) {
channel_map.unlock();
return 2;
}
std::string mi_host{mi->host};
uint mi_port = mi->port;
const std::string mi_network_namespace(mi->network_namespace_str());
THD_STAGE_INFO(thd, stage_connecting_to_source);
Mysql_connection *conn =
new Mysql_connection(thd, mi, host, port, mi_network_namespace);
if (!conn->is_connected()) {
LogErr(WARNING_LEVEL, ER_RPL_ASYNC_CHANNEL_CANT_CONNECT, host.c_str(),
port, "", channel.c_str());
delete conn;
conn = nullptr;
channel_map.unlock();
continue;
}
/*
3.3. Get group membership details for ONLINE, RECOVERING, UNREACHABLE
members.
*/
THD_STAGE_INFO(thd, stage_rpl_failover_fetching_source_member_details);
std::vector<RPL_FAILOVER_SOURCE_TUPLE> group_membership_list{};
int err{0};
bool conn_member_needs_to_change{false}, conn_member_quorum_lost{false};
/* Connection details of source who lost majority. It will be used to log */
std::tuple<std::string, std::string, uint>
conn_member_quorum_lost_details{};
std::tie(err, conn_member_needs_to_change, conn_member_quorum_lost,
conn_member_quorum_lost_details) =
get_online_members(thd, mi, conn, failover_table_detail,
group_membership_list, curr_highest_group_weight,
curr_conn_weight);
delete conn;
conn = nullptr;
channel_map.unlock();
if (is_monitor_killed(thd, nullptr)) return 1;
if (err == ER_RPL_ASYNC_GET_GROUP_MEMBERSHIP_DETAILS_ERROR ||
err == ER_RPL_ASYNC_MONITOR_IO_THD_FETCH_GROUP_MAJORITY_ERROR) {
continue;
}
/*
3.4. Store gathered membership details to
replication_asynchronous_connection_failover table.
*/
THD_STAGE_INFO(thd, stage_rpl_failover_updating_source_member_details);
if (!err && !group_membership_list.empty() &&
!save_group_members(channel, managed_name, group_membership_list)) {
/*
Add the managed_name to the managed_name_list so that further members
from the group can be ignored.
*/
managed_name_list.insert(managed_name);
} else if (err == 2) {
return 1;
}
if (is_monitor_killed(thd, nullptr)) return 1;
/*
3.5. Disconnect channel if current connected member through
asynchronous channel has changed the group or has lost quorum.
*/
if (conn_member_needs_to_change || conn_member_quorum_lost) {
const std::string error_channel =
std::get<0>(conn_member_quorum_lost_details);
const std::string error_host =
std::get<1>(conn_member_quorum_lost_details);
const uint error_port = std::get<2>(conn_member_quorum_lost_details);
/* Get current values from mi. */
channel_map.rdlock();
mi = channel_map.get_mi(channel_name.c_str());
if (nullptr == mi) {
channel_map.unlock();
return 2;
}
mi_host.assign(mi->host);
mi_port = mi->port;
channel_map.unlock();
/*
Only trigger the channel reconnection if the sender on which we detect
the error is still the connected sender.
Until we reached this point the IO thread may had switch by itself to
another sender.
*/
if (!error_channel.compare(channel_name) &&
!error_host.compare(mi_host) && error_port == mi_port) {
if (is_monitor_killed(thd, nullptr)) return 1;
bool restarted = restart_io_thread(thd, channel_name, false);
if (restarted && conn_member_quorum_lost) {
LogErr(ERROR_LEVEL, ER_RPL_ASYNC_CHANNEL_STOPPED_QUORUM_LOST,
error_host.c_str(), error_port, "", error_channel.c_str());
}
}
}
}
if (is_monitor_killed(thd, nullptr)) return 1;
/*
4. Get highest weight of single sender.
*/
{
channel_map.rdlock();
Master_info *mi = channel_map.get_mi(channel_name.c_str());
if (nullptr == mi) {
channel_map.unlock();
return 2;
}
for (auto source_conn_detail : source_conn_detail_list) {
std::string group_name{""};
uint weight{0};
/* save highest weight of single senders for the current channel */
std::tie(std::ignore, std::ignore, std::ignore, std::ignore, weight,
group_name) = source_conn_detail;
if (weight > curr_highest_weight_single_sender && group_name.empty() &&
weight > curr_conn_weight && weight > curr_highest_group_weight &&
check_connection_and_run_query(thd, mi, source_conn_detail)) {
curr_highest_weight_single_sender = weight;
}
}
channel_map.unlock();
}
if (is_monitor_killed(thd, nullptr)) return 1;
/*
5. If weight of current connected sender is less then any of
ONLINE group member or single server, then disconnect it.
The reconnection would be done by IO thread.
*/
DBUG_EXECUTE_IF("async_conn_failover_disable_weight_check", return 0;);
DBUG_EXECUTE_IF("async_conn_failover_check_interim_sender", {
if (source_conn_detail_list.size() == 4) {
return 0;
}
});
if ((curr_highest_group_weight > curr_conn_weight) ||
(curr_highest_weight_single_sender > curr_conn_weight)) {
restart_io_thread(thd, channel_name, true);
}
return 0;
}
bool Source_IO_monitor::check_connection_and_run_query(
THD *thd, Master_info *mi, RPL_FAILOVER_SOURCE_TUPLE &conn_detail) {
uint query_failed{1};
uint port{0};
std::string host{""};
const std::string mi_network_namespace(mi->network_namespace_str());
std::tie(std::ignore, host, port, std::ignore, std::ignore, std::ignore) =
conn_detail;
Mysql_connection *conn_single_server =
new Mysql_connection(thd, mi, host, port, mi_network_namespace);
if (conn_single_server != nullptr && conn_single_server->is_connected())
std::tie(query_failed, std::ignore) = execute_query(
conn_single_server, enum_sql_query_tag::QUERY_SERVER_SELECT_ONE);
if (query_failed != 0 && conn_single_server != nullptr) {
Async_conn_failover_manager::log_error_for_async_executing_query_failure(
ER_RPL_ASYNC_CHECK_CONNECTION_ERROR, conn_single_server->get_mysql(),
mi);
}
delete conn_single_server;
conn_single_server = nullptr;
return !query_failed;
}
int Source_IO_monitor::save_group_members(
std::string channel_name, std::string managed_name,
std::vector<RPL_FAILOVER_SOURCE_TUPLE> &group_membership_list) {
bool err_val{true};
std::string err_msg{};
std::vector<RPL_FAILOVER_SOURCE_TUPLE> failover_table_detail_list{};
std::string db{"mysql"};
std::string table_name{"replication_asynchronous_connection_failover"};
uint num_field{6};
enum thr_lock_type lock_type { TL_WRITE };
std::vector<std::string> field_name{
"channel", "host", "port", "network_namespace", "weight", "managed_name"};
/* Open table with OPTION_AUTOCOMMIT disable. */
Rpl_sys_table_access table_op(db, table_name, num_field);
if (table_op.open(lock_type)) {
table_op.set_error();
return 1;
}
TABLE *table{table_op.get_table()};
/*
Read stored source details from
replication_asynchronous_connection_failover table.
*/
{
/* Store channel */
if (table_op.store_field(table->field[0], channel_name)) {
table_op.set_error();
return 1;
}
/* Store managed_name */
if (table_op.store_field(table->field[5], managed_name)) {
table_op.set_error();
}
Rpl_sys_key_access key_access;
if (!key_access.init(table, 1, true, (key_part_map)((1L << 0) | (1L << 1)),
HA_READ_KEY_EXACT)) {
do {
/* get source detail */
RPL_FAILOVER_SOURCE_TUPLE source_tuple{};
Rpl_async_conn_failover_table_operations::get_data<
RPL_FAILOVER_SOURCE_TUPLE>(table_op, source_tuple);
failover_table_detail_list.push_back(source_tuple);
} while (!key_access.next());
}
if (key_access.deinit()) {
table_op.set_error();
return 1;
}
if (failover_table_detail_list.empty()) return 1;
}
/*
For each source from gathered membership details in step 3,
check whether it's already present in failover table:
- if present then delete its entry and insert again (weight can change).
- if not present then insert.
*/
for (auto group_member_detail : group_membership_list) {
uint port{0}, weight{0};
std::string channel{}, host{}, group_name{}, net_ns{};
std::tie(channel, host, port, net_ns, weight, group_name) =
group_member_detail;
auto it = std::find_if(failover_table_detail_list.begin(),
failover_table_detail_list.end(),
[&](const RPL_FAILOVER_SOURCE_TUPLE &e) {
return ((std::get<0>(e).compare(channel) == 0) &&
(std::get<1>(e).compare(host) == 0) &&
(std::get<2>(e) == port));
});
if (it != failover_table_detail_list.end()) {
auto del_conn_detail = std::make_tuple(channel, host, port);
std::tie(err_val, err_msg) =
delete_rows(table_op, table, field_name, del_conn_detail);
if (err_val) {
table_op.set_error();
table_op.close(err_val);
return 1;
}
std::tie(err_val, err_msg) =
write_rows(table_op, table, field_name, group_member_detail);
if (err_val) {
table_op.set_error();
table_op.close(err_val);
return 1;
}
}
if (it == failover_table_detail_list.end()) {
LogErr(SYSTEM_LEVEL, ER_RPL_ASYNC_SENDER_ADDED, host.c_str(), port, "",
channel.c_str(), group_name.c_str());
std::tie(err_val, err_msg) =
write_rows(table_op, table, field_name, group_member_detail);
if (err_val) {
table_op.set_error();
table_op.close(err_val);
return 1;
}
}
}
/*
For each source from failover table, check whether it was
also found in membership details list :
- if not found then delete its entry from failover table, as
source has left the group.
*/
for (auto failover_table_detail : failover_table_detail_list) {
uint port{0}, weight{0};
std::string channel{}, host{}, group_name{}, net_ns{};
std::tie(channel, host, port, net_ns, weight, group_name) =
failover_table_detail;
auto it =
std::find_if(group_membership_list.begin(), group_membership_list.end(),
[&](const RPL_FAILOVER_SOURCE_TUPLE &e) {
return ((std::get<1>(e).compare(host) == 0) &&
(std::get<2>(e) == port));
});
if (it == group_membership_list.end()) {
LogErr(SYSTEM_LEVEL, ER_RPL_ASYNC_SENDER_REMOVED, host.c_str(), port, "",
channel.c_str(), group_name.c_str());
auto del_conn_detail = std::make_tuple(channel, host, port);
std::tie(err_val, err_msg) =
delete_rows(table_op, table, field_name, del_conn_detail);
if (err_val) {
table_op.set_error();
table_op.close(err_val);
return 1;
}
}
}
/* Increment member action configuration version. */
if (table_op.increment_version()) {
LogErr(ERROR_LEVEL, ER_RPL_INCREMENTING_MEMBER_ACTION_VERSION, db.c_str(),
table_name.c_str());
return 1;
}
/*
Send replication_asynchronous_connection_failover data to group replication
group members.
*/
if (rpl_acf_configuration_handler->send_failover_data(table_op)) {
return 1;
}
return 0;
}
bool Source_IO_monitor::has_primary_lost_contact_with_majority() {
bool primary_lost_contact_with_majority = false;
my_h_service gr_status_service_handler = nullptr;
srv_registry->acquire("group_replication_status_service_v1",
&gr_status_service_handler);
if (nullptr != gr_status_service_handler) {
SERVICE_TYPE(group_replication_status_service_v1) *gr_status_service =
reinterpret_cast<SERVICE_TYPE(group_replication_status_service_v1) *>(
gr_status_service_handler);
if (gr_status_service
->is_group_in_single_primary_mode_and_im_the_primary() &&
!gr_status_service->is_member_online_with_group_majority()) {
primary_lost_contact_with_majority = true;
}
srv_registry->release(gr_status_service_handler);
}
return primary_lost_contact_with_majority;
}
std::tuple<int, bool, bool, std::tuple<std::string, std::string, uint>>
Source_IO_monitor::get_online_members(
THD *thd, Master_info *mi, const Mysql_connection *conn,
SENDER_CONN_MERGE_TUPLE failover_table_detail,
std::vector<RPL_FAILOVER_SOURCE_TUPLE> &group_membership_list,
uint &curr_highest_group_weight, uint &curr_conn_weight) {
channel_map.assert_some_lock();
uint error{0};
std::string err_msg;
bool conn_member_needs_to_change{false}, conn_member_quorum_lost{false};
/* Connection details of source who lost majority. It will be used to log */
std::tuple<std::string, std::string, uint> conn_member_quorum_lost_details{};
uint primary_weight{0}, secondary_weight{0}, port{0}, weight{0};
std::string channel{}, host{}, managed_name{};
if (is_monitor_killed(thd, nullptr)) {
return std::make_tuple(2, conn_member_needs_to_change,
conn_member_quorum_lost,
conn_member_quorum_lost_details);
}
std::tie(channel, host, port, std::ignore, weight, managed_name,
primary_weight, secondary_weight) = failover_table_detail;
/* Execute enum_sql_query_tag::CONFIG_MODE_QUORUM_MONITOR query */
MYSQL_RES_VAL quorum_list{};
auto qtag{enum_sql_query_tag::CONFIG_MODE_QUORUM_MONITOR};
std::tie(error, quorum_list) = execute_query(conn, qtag);
if (error != 0) {
longlong sql_errno{ER_RPL_ASYNC_MONITOR_IO_THD_FETCH_GROUP_MAJORITY_ERROR};
Async_conn_failover_manager::log_error_for_async_executing_query_failure(
sql_errno, const_cast<Mysql_connection *>(conn)->get_mysql(), mi);
return std::make_tuple(sql_errno, conn_member_needs_to_change,
conn_member_quorum_lost,
conn_member_quorum_lost_details);
}
if (quorum_list.empty() || quorum_list[0].empty()) {
return std::make_tuple(1, conn_member_needs_to_change,
conn_member_quorum_lost,
conn_member_quorum_lost_details);
}
auto quorum_status{
static_cast<enum_conf_mode_quorum_status>(std::stoi(quorum_list[0][0]))};
if (quorum_status == enum_conf_mode_quorum_status::MANAGED_GR_HAS_QUORUM) {
qtag = enum_sql_query_tag::GR_MEMBER_ALL_DETAILS;
MYSQL_RES_VAL sender_membership_res{};
std::tie(error, sender_membership_res) = execute_query(conn, qtag);
if (error == ER_BAD_FIELD_ERROR) {
qtag = enum_sql_query_tag::GR_MEMBER_ALL_DETAILS_FETCH_FOR_57;
std::tie(error, sender_membership_res) = execute_query(conn, qtag);
}
if (error != 0) {
longlong sql_errno{ER_RPL_ASYNC_GET_GROUP_MEMBERSHIP_DETAILS_ERROR};
Async_conn_failover_manager::log_error_for_async_executing_query_failure(
sql_errno, const_cast<Mysql_connection *>(conn)->get_mysql(), mi);
return std::make_tuple(sql_errno, conn_member_needs_to_change,
conn_member_quorum_lost,
conn_member_quorum_lost_details);
}
/*
If current connected sender is group member and not a single server
then save its primary/secondary weight based on role.
*/
for (const auto &m_row_ins : sender_membership_res) {
if (m_row_ins[COL_HOST].compare(mi->host) == 0 &&
std::stoul(m_row_ins[COL_PORT]) == mi->port) {
if (m_row_ins[COL_ROLE].compare("PRIMARY") == 0) {
curr_conn_weight = primary_weight;
} else if (m_row_ins[COL_ROLE].compare("SECONDARY") == 0) {
curr_conn_weight = secondary_weight;
}
}
}
for (const auto &m_row : sender_membership_res) {
/*
If member is ONLINE then add member connection details to
replication_asynchronous_connection_failover table.
*/
if (m_row[COL_STATE].compare("ONLINE") == 0 ||
m_row[COL_STATE].compare("RECOVERING") == 0 ||
m_row[COL_STATE].compare("UNREACHABLE") == 0) {
if (is_monitor_killed(thd, nullptr)) {
return std::make_tuple(2, conn_member_needs_to_change,
conn_member_quorum_lost,
conn_member_quorum_lost_details);
}
uint tab_weight{secondary_weight};
if (m_row[COL_ROLE].compare("PRIMARY") == 0) {
tab_weight = primary_weight;
if ((primary_weight > curr_highest_group_weight) &&
m_row[COL_STATE].compare("ONLINE") == 0)
curr_highest_group_weight = primary_weight;
} else if (m_row[COL_ROLE].compare("SECONDARY") == 0) {
if ((secondary_weight > curr_highest_group_weight) &&
m_row[COL_STATE].compare("ONLINE") == 0)
curr_highest_group_weight = secondary_weight;
}
auto source_ins_details = std::make_tuple(
channel, m_row[COL_HOST], std::stoul(m_row[COL_PORT]), "",
tab_weight, m_row[COL_GROUP_NAME]);
group_membership_list.push_back(source_ins_details);
}
/*
For the source connected through asynchronous channel,
if the group_name has changed i.e. member has changed group, or,
if its state become UNREACHABLE i.e. lost majority, then stop
the asynchronous channel.
*/
if (m_row[COL_HOST].compare(mi->host) == 0 &&
std::stoul(m_row[COL_PORT]) == mi->port &&
(m_row[COL_GROUP_NAME].compare(managed_name) != 0 ||
m_row[COL_STATE].compare("UNREACHABLE") == 0)) {
conn_member_needs_to_change = true;
}
}
}
if (quorum_status == enum_conf_mode_quorum_status::MANAGED_GR_HAS_ERROR &&
host.compare(mi->host) == 0 && port == mi->port) {
conn_member_quorum_lost = true;
conn_member_quorum_lost_details = std::make_tuple(channel, host, port);
}
return std::make_tuple(0, conn_member_needs_to_change,
conn_member_quorum_lost,
conn_member_quorum_lost_details);
}
int Source_IO_monitor::sync_senders_details(THD *thd) {
bool primary_lost_contact_with_majority =
has_primary_lost_contact_with_majority();
if (primary_lost_contact_with_majority) {
/* Log the warning only once per majority loss. */
if (!m_primary_lost_contact_with_majority_warning_logged) {
m_primary_lost_contact_with_majority_warning_logged = true;
LogErr(WARNING_LEVEL, ER_GRP_RPL_FAILOVER_PRIMARY_WITHOUT_MAJORITY);
}
return 0;
} else {
if (m_primary_lost_contact_with_majority_warning_logged) {
m_primary_lost_contact_with_majority_warning_logged = false;
LogErr(WARNING_LEVEL, ER_GRP_RPL_FAILOVER_PRIMARY_BACK_TO_MAJORITY);
}
}
std::vector<std::string> channels;
channel_map.rdlock();
for (mi_map::iterator it = channel_map.begin(); it != channel_map.end();
it++) {
Master_info *mi = it->second;
if (Master_info::is_configured(mi) &&
mi->is_source_connection_auto_failover()) {
channels.push_back(mi->get_channel());
}
}
channel_map.unlock();
for (const std::string &channel_name : channels) {
connect_senders(thd, channel_name);
}
return 0;
}
std::tuple<bool, std::vector<SENDER_CONN_MERGE_TUPLE>>
Source_IO_monitor::get_senders_details(const std::string &channel_name) {
DBUG_TRACE;
/* The list of source connection details. */
std::vector<SENDER_CONN_MERGE_TUPLE> failover_table_detail_list{};
std::vector<RPL_FAILOVER_MANAGED_TUPLE> source_managed_list{};
auto error{false};
/*
Check if source needs to be managed, if true then get its network
configuration details.
These tables can be modified in parallel, which will cause its open() to
fail, on that case we do retry the operation.
*/
int retries = 0;
do {
if (retries > 0) {
my_sleep(1000);
}
Rpl_async_conn_failover_table_operations table_op(TL_READ);
error = table_op.read_managed_rows_for_channel(channel_name,
source_managed_list);
if (error) return make_pair(error, failover_table_detail_list);