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 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514
|
// repl.cpp
/* TODO
PAIRING
_ on a syncexception, don't allow going back to master state?
*/
/**
* Copyright (C) 2008 10gen Inc.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License, version 3,
* as published by the Free Software Foundation.
*
* 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 Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
/* Collections we use:
local.sources - indicates what sources we pull from as a "slave", and the last update of each
local.oplog.$main - our op log as "master"
local.dbinfo.<dbname> - no longer used???
local.pair.startup - [deprecated] can contain a special value indicating for a pair that we have the master copy.
used when replacing other half of the pair which has permanently failed.
local.pair.sync - [deprecated] { initialsynccomplete: 1 }
*/
#include "pch.h"
#include "jsobj.h"
#include "../util/goodies.h"
#include "repl.h"
#include "../util/net/message.h"
#include "../util/background.h"
#include "../client/dbclient.h"
#include "../client/connpool.h"
#include "pdfile.h"
#include "ops/query.h"
#include "db.h"
#include "commands.h"
#include "security.h"
#include "cmdline.h"
#include "repl_block.h"
#include "repl/rs.h"
#include "replutil.h"
#include "repl/connections.h"
#include "ops/update.h"
namespace mongo {
// our config from command line etc.
ReplSettings replSettings;
/* if 1 sync() is running */
volatile int syncing = 0;
static volatile int relinquishSyncingSome = 0;
/* "dead" means something really bad happened like replication falling completely out of sync.
when non-null, we are dead and the string is informational
*/
const char *replAllDead = 0;
time_t lastForcedResync = 0;
} // namespace mongo
namespace mongo {
/* output by the web console */
const char *replInfo = "";
struct ReplInfo {
ReplInfo(const char *msg) {
replInfo = msg;
}
~ReplInfo() {
replInfo = "?";
}
};
/* operator requested resynchronization of replication (on the slave). { resync : 1 } */
class CmdResync : public Command {
public:
virtual bool slaveOk() const {
return true;
}
virtual bool adminOnly() const {
return true;
}
virtual bool logTheOp() { return false; }
virtual LockType locktype() const { return WRITE; }
void help(stringstream&h) const { h << "resync (from scratch) an out of date replica slave.\nhttp://www.mongodb.org/display/DOCS/Master+Slave"; }
CmdResync() : Command("resync") { }
virtual bool run(const string& , BSONObj& cmdObj, int, string& errmsg, BSONObjBuilder& result, bool fromRepl) {
if( cmdLine.usingReplSets() ) {
errmsg = "resync command not currently supported with replica sets. See RS102 info in the mongodb documentations";
result.append("info", "http://www.mongodb.org/display/DOCS/Resyncing+a+Very+Stale+Replica+Set+Member");
return false;
}
if ( cmdObj.getBoolField( "force" ) ) {
if ( !waitForSyncToFinish( errmsg ) )
return false;
replAllDead = "resync forced";
}
if ( !replAllDead ) {
errmsg = "not dead, no need to resync";
return false;
}
if ( !waitForSyncToFinish( errmsg ) )
return false;
ReplSource::forceResyncDead( "client" );
result.append( "info", "triggered resync for all sources" );
return true;
}
bool waitForSyncToFinish( string &errmsg ) const {
// Wait for slave thread to finish syncing, so sources will be be
// reloaded with new saved state on next pass.
Timer t;
while ( 1 ) {
if ( syncing == 0 || t.millis() > 30000 )
break;
{
dbtemprelease t;
relinquishSyncingSome = 1;
sleepmillis(1);
}
}
if ( syncing ) {
errmsg = "timeout waiting for sync() to finish";
return false;
}
return true;
}
} cmdResync;
bool anyReplEnabled() {
return replSettings.slave || replSettings.master || theReplSet;
}
bool replAuthenticate(DBClientBase *conn);
void appendReplicationInfo( BSONObjBuilder& result , bool authed , int level ) {
if ( replSet ) {
if( theReplSet == 0 ) {
result.append("ismaster", false);
result.append("secondary", false);
result.append("info", ReplSet::startupStatusMsg.get());
result.append( "isreplicaset" , true );
return;
}
theReplSet->fillIsMaster(result);
return;
}
if ( replAllDead ) {
result.append("ismaster", 0);
string s = string("dead: ") + replAllDead;
result.append("info", s);
}
else {
result.appendBool("ismaster", _isMaster() );
}
if ( level && replSet ) {
result.append( "info" , "is replica set" );
}
else if ( level ) {
BSONObjBuilder sources( result.subarrayStart( "sources" ) );
readlock lk( "local.sources" );
Client::Context ctx( "local.sources", dbpath, 0, authed );
shared_ptr<Cursor> c = findTableScan("local.sources", BSONObj());
int n = 0;
while ( c->ok() ) {
BSONObj s = c->current();
BSONObjBuilder bb;
bb.append( s["host"] );
string sourcename = s["source"].valuestr();
if ( sourcename != "main" )
bb.append( s["source"] );
{
BSONElement e = s["syncedTo"];
BSONObjBuilder t( bb.subobjStart( "syncedTo" ) );
t.appendDate( "time" , e.timestampTime() );
t.append( "inc" , e.timestampInc() );
t.done();
}
if ( level > 1 ) {
dbtemprelease unlock;
// note: there is no so-style timeout on this connection; perhaps we should have one.
ScopedDbConnection conn( s["host"].valuestr() );
DBClientConnection *cliConn = dynamic_cast< DBClientConnection* >( &conn.conn() );
if ( cliConn && replAuthenticate( cliConn ) ) {
BSONObj first = conn->findOne( (string)"local.oplog.$" + sourcename , Query().sort( BSON( "$natural" << 1 ) ) );
BSONObj last = conn->findOne( (string)"local.oplog.$" + sourcename , Query().sort( BSON( "$natural" << -1 ) ) );
bb.appendDate( "masterFirst" , first["ts"].timestampTime() );
bb.appendDate( "masterLast" , last["ts"].timestampTime() );
double lag = (double) (last["ts"].timestampTime() - s["syncedTo"].timestampTime());
bb.append( "lagSeconds" , lag / 1000 );
}
conn.done();
}
sources.append( BSONObjBuilder::numStr( n++ ) , bb.obj() );
c->advance();
}
sources.done();
}
}
class CmdIsMaster : public Command {
public:
virtual bool requiresAuth() { return false; }
virtual bool slaveOk() const {
return true;
}
virtual void help( stringstream &help ) const {
help << "Check if this server is primary for a replica pair/set; also if it is --master or --slave in simple master/slave setups.\n";
help << "{ isMaster : 1 }";
}
virtual LockType locktype() const { return NONE; }
CmdIsMaster() : Command("isMaster", true, "ismaster") { }
virtual bool run(const string& , BSONObj& cmdObj, int, string& errmsg, BSONObjBuilder& result, bool /*fromRepl*/) {
/* currently request to arbiter is (somewhat arbitrarily) an ismaster request that is not
authenticated.
we allow unauthenticated ismaster but we aren't as verbose informationally if
one is not authenticated for admin db to be safe.
*/
bool authed = cc().getAuthenticationInfo()->isAuthorizedReads("admin");
appendReplicationInfo( result , authed );
result.appendNumber("maxBsonObjectSize", BSONObjMaxUserSize);
return true;
}
} cmdismaster;
ReplSource::ReplSource() {
nClonedThisPass = 0;
}
ReplSource::ReplSource(BSONObj o) : nClonedThisPass(0) {
only = o.getStringField("only");
hostName = o.getStringField("host");
_sourceName = o.getStringField("source");
uassert( 10118 , "'host' field not set in sources collection object", !hostName.empty() );
uassert( 10119 , "only source='main' allowed for now with replication", sourceName() == "main" );
BSONElement e = o.getField("syncedTo");
if ( !e.eoo() ) {
uassert( 10120 , "bad sources 'syncedTo' field value", e.type() == Date || e.type() == Timestamp );
OpTime tmp( e.date() );
syncedTo = tmp;
}
BSONObj dbsObj = o.getObjectField("dbsNextPass");
if ( !dbsObj.isEmpty() ) {
BSONObjIterator i(dbsObj);
while ( 1 ) {
BSONElement e = i.next();
if ( e.eoo() )
break;
addDbNextPass.insert( e.fieldName() );
}
}
dbsObj = o.getObjectField("incompleteCloneDbs");
if ( !dbsObj.isEmpty() ) {
BSONObjIterator i(dbsObj);
while ( 1 ) {
BSONElement e = i.next();
if ( e.eoo() )
break;
incompleteCloneDbs.insert( e.fieldName() );
}
}
}
/* Turn our C++ Source object into a BSONObj */
BSONObj ReplSource::jsobj() {
BSONObjBuilder b;
b.append("host", hostName);
b.append("source", sourceName());
if ( !only.empty() )
b.append("only", only);
if ( !syncedTo.isNull() )
b.appendTimestamp("syncedTo", syncedTo.asDate());
BSONObjBuilder dbsNextPassBuilder;
int n = 0;
for ( set<string>::iterator i = addDbNextPass.begin(); i != addDbNextPass.end(); i++ ) {
n++;
dbsNextPassBuilder.appendBool(*i, 1);
}
if ( n )
b.append("dbsNextPass", dbsNextPassBuilder.done());
BSONObjBuilder incompleteCloneDbsBuilder;
n = 0;
for ( set<string>::iterator i = incompleteCloneDbs.begin(); i != incompleteCloneDbs.end(); i++ ) {
n++;
incompleteCloneDbsBuilder.appendBool(*i, 1);
}
if ( n )
b.append("incompleteCloneDbs", incompleteCloneDbsBuilder.done());
return b.obj();
}
void ReplSource::save() {
BSONObjBuilder b;
assert( !hostName.empty() );
b.append("host", hostName);
// todo: finish allowing multiple source configs.
// this line doesn't work right when source is null, if that is allowed as it is now:
//b.append("source", _sourceName);
BSONObj pattern = b.done();
BSONObj o = jsobj();
log( 1 ) << "Saving repl source: " << o << endl;
{
OpDebug debug;
Client::Context ctx("local.sources");
UpdateResult res = updateObjects("local.sources", o, pattern, true/*upsert for pair feature*/, false,false,debug);
assert( ! res.mod );
assert( res.num == 1 );
}
}
static void addSourceToList(ReplSource::SourceVector &v, ReplSource& s, ReplSource::SourceVector &old) {
if ( !s.syncedTo.isNull() ) { // Don't reuse old ReplSource if there was a forced resync.
for ( ReplSource::SourceVector::iterator i = old.begin(); i != old.end(); ) {
if ( s == **i ) {
v.push_back(*i);
old.erase(i);
return;
}
i++;
}
}
v.push_back( shared_ptr< ReplSource >( new ReplSource( s ) ) );
}
/* we reuse our existing objects so that we can keep our existing connection
and cursor in effect.
*/
void ReplSource::loadAll(SourceVector &v) {
Client::Context ctx("local.sources");
SourceVector old = v;
v.clear();
if ( !cmdLine.source.empty() ) {
// --source <host> specified.
// check that no items are in sources other than that
// add if missing
shared_ptr<Cursor> c = findTableScan("local.sources", BSONObj());
int n = 0;
while ( c->ok() ) {
n++;
ReplSource tmp(c->current());
if ( tmp.hostName != cmdLine.source ) {
log() << "repl: --source " << cmdLine.source << " != " << tmp.hostName << " from local.sources collection" << endl;
log() << "repl: for instructions on changing this slave's source, see:" << endl;
log() << "http://dochub.mongodb.org/core/masterslave" << endl;
log() << "repl: terminating mongod after 30 seconds" << endl;
sleepsecs(30);
dbexit( EXIT_REPLICATION_ERROR );
}
if ( tmp.only != cmdLine.only ) {
log() << "--only " << cmdLine.only << " != " << tmp.only << " from local.sources collection" << endl;
log() << "terminating after 30 seconds" << endl;
sleepsecs(30);
dbexit( EXIT_REPLICATION_ERROR );
}
c->advance();
}
uassert( 10002 , "local.sources collection corrupt?", n<2 );
if ( n == 0 ) {
// source missing. add.
ReplSource s;
s.hostName = cmdLine.source;
s.only = cmdLine.only;
s.save();
}
}
else {
try {
massert( 10384 , "--only requires use of --source", cmdLine.only.empty());
}
catch ( ... ) {
dbexit( EXIT_BADOPTIONS );
}
}
shared_ptr<Cursor> c = findTableScan("local.sources", BSONObj());
while ( c->ok() ) {
ReplSource tmp(c->current());
if ( tmp.syncedTo.isNull() ) {
DBDirectClient c;
if ( c.exists( "local.oplog.$main" ) ) {
BSONObj op = c.findOne( "local.oplog.$main", QUERY( "op" << NE << "n" ).sort( BSON( "$natural" << -1 ) ) );
if ( !op.isEmpty() ) {
tmp.syncedTo = op[ "ts" ].date();
}
}
}
addSourceToList(v, tmp, old);
c->advance();
}
}
BSONObj opTimeQuery = fromjson("{\"getoptime\":1}");
bool ReplSource::throttledForceResyncDead( const char *requester ) {
if ( time( 0 ) - lastForcedResync > 600 ) {
forceResyncDead( requester );
lastForcedResync = time( 0 );
return true;
}
return false;
}
void ReplSource::forceResyncDead( const char *requester ) {
if ( !replAllDead )
return;
SourceVector sources;
ReplSource::loadAll(sources);
for( SourceVector::iterator i = sources.begin(); i != sources.end(); ++i ) {
log() << requester << " forcing resync from " << (*i)->hostName << endl;
(*i)->forceResync( requester );
}
replAllDead = 0;
}
void ReplSource::forceResync( const char *requester ) {
BSONObj info;
{
dbtemprelease t;
if (!oplogReader.connect(hostName)) {
msgassertedNoTrace( 14051 , "unable to connect to resync");
}
/* todo use getDatabaseNames() method here */
bool ok = oplogReader.conn()->runCommand( "admin", BSON( "listDatabases" << 1 ), info );
massert( 10385 , "Unable to get database list", ok );
}
BSONObjIterator i( info.getField( "databases" ).embeddedObject() );
while( i.moreWithEOO() ) {
BSONElement e = i.next();
if ( e.eoo() )
break;
string name = e.embeddedObject().getField( "name" ).valuestr();
if ( !e.embeddedObject().getBoolField( "empty" ) ) {
if ( name != "local" ) {
if ( only.empty() || only == name ) {
resyncDrop( name.c_str(), requester );
}
}
}
}
syncedTo = OpTime();
addDbNextPass.clear();
save();
}
string ReplSource::resyncDrop( const char *db, const char *requester ) {
log() << "resync: dropping database " << db << endl;
Client::Context ctx(db);
dropDatabase(db);
return db;
}
/* grab initial copy of a database from the master */
void ReplSource::resync(string db) {
string dummyNs = resyncDrop( db.c_str(), "internal" );
Client::Context ctx( dummyNs );
{
log() << "resync: cloning database " << db << " to get an initial copy" << endl;
ReplInfo r("resync: cloning a database");
string errmsg;
int errCode = 0;
bool ok = cloneFrom(hostName.c_str(), errmsg, cc().database()->name, false, /*slaveok*/ true, /*replauth*/ true, /*snapshot*/false, /*mayYield*/true, /*mayBeInterrupted*/false, &errCode);
if ( !ok ) {
if ( errCode == DatabaseDifferCaseCode ) {
resyncDrop( db.c_str(), "internal" );
log() << "resync: database " << db << " not valid on the master due to a name conflict, dropping." << endl;
return;
}
else {
problem() << "resync of " << db << " from " << hostName << " failed " << errmsg << endl;
throw SyncException();
}
}
}
log() << "resync: done with initial clone for db: " << db << endl;
return;
}
DatabaseIgnorer ___databaseIgnorer;
void DatabaseIgnorer::doIgnoreUntilAfter( const string &db, const OpTime &futureOplogTime ) {
if ( futureOplogTime > _ignores[ db ] ) {
_ignores[ db ] = futureOplogTime;
}
}
bool DatabaseIgnorer::ignoreAt( const string &db, const OpTime ¤tOplogTime ) {
if ( _ignores[ db ].isNull() ) {
return false;
}
if ( _ignores[ db ] >= currentOplogTime ) {
return true;
} else {
// The ignore state has expired, so clear it.
_ignores.erase( db );
return false;
}
}
bool ReplSource::handleDuplicateDbName( const BSONObj &op, const char *ns, const char *db ) {
if ( dbHolder.isLoaded( ns, dbpath ) ) {
// Database is already present.
return true;
}
BSONElement ts = op.getField( "ts" );
if ( ( ts.type() == Date || ts.type() == Timestamp ) && ___databaseIgnorer.ignoreAt( db, ts.date() ) ) {
// Database is ignored due to a previous indication that it is
// missing from master after optime "ts".
return false;
}
if ( Database::duplicateUncasedName( db, dbpath ).empty() ) {
// No duplicate database names are present.
return true;
}
OpTime lastTime;
bool dbOk = false;
{
dbtemprelease release;
// We always log an operation after executing it (never before), so
// a database list will always be valid as of an oplog entry generated
// before it was retrieved.
BSONObj last = oplogReader.findOne( this->ns().c_str(), Query().sort( BSON( "$natural" << -1 ) ) );
if ( !last.isEmpty() ) {
BSONElement ts = last.getField( "ts" );
massert( 14032, "Invalid 'ts' in remote log", ts.type() == Date || ts.type() == Timestamp );
lastTime = OpTime( ts.date() );
}
BSONObj info;
bool ok = oplogReader.conn()->runCommand( "admin", BSON( "listDatabases" << 1 ), info );
massert( 14033, "Unable to get database list", ok );
BSONObjIterator i( info.getField( "databases" ).embeddedObject() );
while( i.more() ) {
BSONElement e = i.next();
const char * name = e.embeddedObject().getField( "name" ).valuestr();
if ( strcasecmp( name, db ) != 0 )
continue;
if ( strcmp( name, db ) == 0 ) {
// The db exists on master, still need to check that no conflicts exist there.
dbOk = true;
continue;
}
// The master has a db name that conflicts with the requested name.
dbOk = false;
break;
}
}
if ( !dbOk ) {
___databaseIgnorer.doIgnoreUntilAfter( db, lastTime );
incompleteCloneDbs.erase(db);
addDbNextPass.erase(db);
return false;
}
// Check for duplicates again, since we released the lock above.
set< string > duplicates;
Database::duplicateUncasedName( db, dbpath, &duplicates );
// The database is present on the master and no conflicting databases
// are present on the master. Drop any local conflicts.
for( set< string >::const_iterator i = duplicates.begin(); i != duplicates.end(); ++i ) {
___databaseIgnorer.doIgnoreUntilAfter( *i, lastTime );
incompleteCloneDbs.erase(*i);
addDbNextPass.erase(*i);
Client::Context ctx(*i);
dropDatabase(*i);
}
massert( 14034, "Duplicate database names present after attempting to delete duplicates",
Database::duplicateUncasedName( db, dbpath ).empty() );
return true;
}
void ReplSource::applyOperation(const BSONObj& op) {
try {
bool failedUpdate = applyOperation_inlock( op );
if (failedUpdate && shouldRetry(op, hostName)) {
failedUpdate = applyOperation_inlock( op );
uassert(15914, "Failure retrying initial sync update", ! failedUpdate );
}
}
catch ( UserException& e ) {
log() << "sync: caught user assertion " << e << " while applying op: " << op << endl;;
}
catch ( DBException& e ) {
log() << "sync: caught db exception " << e << " while applying op: " << op << endl;;
}
}
/* local.$oplog.main is of the form:
{ ts: ..., op: <optype>, ns: ..., o: <obj> , o2: <extraobj>, b: <boolflag> }
...
see logOp() comments.
@param alreadyLocked caller already put us in write lock if true
*/
void ReplSource::sync_pullOpLog_applyOperation(BSONObj& op, bool alreadyLocked) {
if( logLevel >= 6 ) // op.tostring is expensive so doing this check explicitly
log(6) << "processing op: " << op << endl;
if( op.getStringField("op")[0] == 'n' )
return;
char clientName[MaxDatabaseNameLen];
const char *ns = op.getStringField("ns");
nsToDatabase(ns, clientName);
if ( *ns == '.' ) {
problem() << "skipping bad op in oplog: " << op.toString() << endl;
return;
}
else if ( *ns == 0 ) {
/*if( op.getStringField("op")[0] != 'n' )*/ {
problem() << "halting replication, bad op in oplog:\n " << op.toString() << endl;
replAllDead = "bad object in oplog";
throw SyncException();
}
//ns = "local.system.x";
//nsToDatabase(ns, clientName);
}
if ( !only.empty() && only != clientName )
return;
if( cmdLine.pretouch && !alreadyLocked/*doesn't make sense if in write lock already*/ ) {
if( cmdLine.pretouch > 1 ) {
/* note: this is bad - should be put in ReplSource. but this is first test... */
static int countdown;
assert( countdown >= 0 );
if( countdown > 0 ) {
countdown--; // was pretouched on a prev pass
}
else {
const int m = 4;
if( tp.get() == 0 ) {
int nthr = min(8, cmdLine.pretouch);
nthr = max(nthr, 1);
tp.reset( new ThreadPool(nthr) );
}
vector<BSONObj> v;
oplogReader.peek(v, cmdLine.pretouch);
unsigned a = 0;
while( 1 ) {
if( a >= v.size() ) break;
unsigned b = a + m - 1; // v[a..b]
if( b >= v.size() ) b = v.size() - 1;
tp->schedule(pretouchN, v, a, b);
DEV cout << "pretouch task: " << a << ".." << b << endl;
a += m;
}
// we do one too...
pretouchOperation(op);
tp->join();
countdown = v.size();
}
}
else {
pretouchOperation(op);
}
}
scoped_ptr<writelock> lk( alreadyLocked ? 0 : new writelock() );
if ( replAllDead ) {
// hmmm why is this check here and not at top of this function? does it get set between top and here?
log() << "replAllDead, throwing SyncException: " << replAllDead << endl;
throw SyncException();
}
if ( !handleDuplicateDbName( op, ns, clientName ) ) {
return;
}
Client::Context ctx( ns );
ctx.getClient()->curop()->reset();
bool empty = ctx.db()->isEmpty();
bool incompleteClone = incompleteCloneDbs.count( clientName ) != 0;
if( logLevel >= 6 )
log(6) << "ns: " << ns << ", justCreated: " << ctx.justCreated() << ", empty: " << empty << ", incompleteClone: " << incompleteClone << endl;
// always apply admin command command
// this is a bit hacky -- the semantics of replication/commands aren't well specified
if ( strcmp( clientName, "admin" ) == 0 && *op.getStringField( "op" ) == 'c' ) {
applyOperation( op );
return;
}
if ( ctx.justCreated() || empty || incompleteClone ) {
// we must add to incomplete list now that setClient has been called
incompleteCloneDbs.insert( clientName );
if ( nClonedThisPass ) {
/* we only clone one database per pass, even if a lot need done. This helps us
avoid overflowing the master's transaction log by doing too much work before going
back to read more transactions. (Imagine a scenario of slave startup where we try to
clone 100 databases in one pass.)
*/
addDbNextPass.insert( clientName );
}
else {
if ( incompleteClone ) {
log() << "An earlier initial clone of '" << clientName << "' did not complete, now resyncing." << endl;
}
save();
Client::Context ctx(ns);
nClonedThisPass++;
resync(ctx.db()->name);
addDbNextPass.erase(clientName);
incompleteCloneDbs.erase( clientName );
}
save();
}
else {
applyOperation( op );
addDbNextPass.erase( clientName );
}
}
void ReplSource::syncToTailOfRemoteLog() {
string _ns = ns();
BSONObjBuilder b;
if ( !only.empty() ) {
b.appendRegex("ns", string("^") + only);
}
BSONObj last = oplogReader.findOne( _ns.c_str(), Query( b.done() ).sort( BSON( "$natural" << -1 ) ) );
if ( !last.isEmpty() ) {
BSONElement ts = last.getField( "ts" );
massert( 10386 , "non Date ts found: " + last.toString(), ts.type() == Date || ts.type() == Timestamp );
syncedTo = OpTime( ts.date() );
}
}
extern unsigned replApplyBatchSize;
/* slave: pull some data from the master's oplog
note: not yet in db mutex at this point.
@return -1 error
0 ok, don't sleep
1 ok, sleep
*/
int ReplSource::sync_pullOpLog(int& nApplied) {
int okResultCode = 1;
string ns = string("local.oplog.$") + sourceName();
log(2) << "repl: sync_pullOpLog " << ns << " syncedTo:" << syncedTo.toStringLong() << '\n';
bool tailing = true;
oplogReader.tailCheck();
bool initial = syncedTo.isNull();
if ( !oplogReader.haveCursor() || initial ) {
if ( initial ) {
// Important to grab last oplog timestamp before listing databases.
syncToTailOfRemoteLog();
BSONObj info;
bool ok = oplogReader.conn()->runCommand( "admin", BSON( "listDatabases" << 1 ), info );
massert( 10389 , "Unable to get database list", ok );
BSONObjIterator i( info.getField( "databases" ).embeddedObject() );
while( i.moreWithEOO() ) {
BSONElement e = i.next();
if ( e.eoo() )
break;
string name = e.embeddedObject().getField( "name" ).valuestr();
if ( !e.embeddedObject().getBoolField( "empty" ) ) {
if ( name != "local" ) {
if ( only.empty() || only == name ) {
log( 2 ) << "adding to 'addDbNextPass': " << name << endl;
addDbNextPass.insert( name );
}
}
}
}
dblock lk;
save();
}
BSONObjBuilder q;
q.appendDate("$gte", syncedTo.asDate());
BSONObjBuilder query;
query.append("ts", q.done());
if ( !only.empty() ) {
// note we may here skip a LOT of data table scanning, a lot of work for the master.
query.appendRegex("ns", string("^") + only); // maybe append "\\." here?
}
BSONObj queryObj = query.done();
// e.g. queryObj = { ts: { $gte: syncedTo } }
oplogReader.tailingQuery(ns.c_str(), queryObj);
tailing = false;
}
else {
log(2) << "repl: tailing=true\n";
}
if( !oplogReader.haveCursor() ) {
problem() << "repl: dbclient::query returns null (conn closed?)" << endl;
oplogReader.resetConnection();
return -1;
}
// show any deferred database creates from a previous pass
{
set<string>::iterator i = addDbNextPass.begin();
if ( i != addDbNextPass.end() ) {
BSONObjBuilder b;
b.append("ns", *i + '.');
b.append("op", "db");
BSONObj op = b.done();
sync_pullOpLog_applyOperation(op, false);
}
}
if ( !oplogReader.more() ) {
if ( tailing ) {
log(2) << "repl: tailing & no new activity\n";
if( oplogReader.awaitCapable() )
okResultCode = 0; // don't sleep
}
else {
log() << "repl: " << ns << " oplog is empty\n";
}
{
dblock lk;
save();
}
return okResultCode;
}
OpTime nextOpTime;
{
BSONObj op = oplogReader.next();
BSONElement ts = op.getField("ts");
if ( ts.type() != Date && ts.type() != Timestamp ) {
string err = op.getStringField("$err");
if ( !err.empty() ) {
// 13051 is "tailable cursor requested on non capped collection"
if (op.getIntField("code") == 13051) {
problem() << "trying to slave off of a non-master" << '\n';
massert( 13344 , "trying to slave off of a non-master", false );
}
else {
problem() << "repl: $err reading remote oplog: " + err << '\n';
massert( 10390 , "got $err reading remote oplog", false );
}
}
else {
problem() << "repl: bad object read from remote oplog: " << op.toString() << '\n';
massert( 10391 , "repl: bad object read from remote oplog", false);
}
}
nextOpTime = OpTime( ts.date() );
log(2) << "repl: first op time received: " << nextOpTime.toString() << '\n';
if ( initial ) {
log(1) << "repl: initial run\n";
}
if( tailing ) {
if( !( syncedTo < nextOpTime ) ) {
log() << "repl ASSERTION failed : syncedTo < nextOpTime" << endl;
log() << "repl syncTo: " << syncedTo.toStringLong() << endl;
log() << "repl nextOpTime: " << nextOpTime.toStringLong() << endl;
assert(false);
}
oplogReader.putBack( op ); // op will be processed in the loop below
nextOpTime = OpTime(); // will reread the op below
}
else if ( nextOpTime != syncedTo ) { // didn't get what we queried for - error
Nullstream& l = log();
l << "repl: nextOpTime " << nextOpTime.toStringLong() << ' ';
if ( nextOpTime < syncedTo )
l << "<??";
else
l << ">";
l << " syncedTo " << syncedTo.toStringLong() << '\n';
log() << "repl: time diff: " << (nextOpTime.getSecs() - syncedTo.getSecs()) << "sec\n";
log() << "repl: tailing: " << tailing << '\n';
log() << "repl: data too stale, halting replication" << endl;
replInfo = replAllDead = "data too stale halted replication";
assert( syncedTo < nextOpTime );
throw SyncException();
}
else {
/* t == syncedTo, so the first op was applied previously or it is the first op of initial query and need not be applied. */
}
}
// apply operations
{
int n = 0;
time_t saveLast = time(0);
while ( 1 ) {
bool moreInitialSyncsPending = !addDbNextPass.empty() && n; // we need "&& n" to assure we actually process at least one op to get a sync point recorded in the first place.
if ( moreInitialSyncsPending || !oplogReader.more() ) {
dblock lk;
// NOTE aaron 2011-03-29 This block may be unnecessary, but I'm leaving it in place to avoid changing timing behavior.
{
dbtemprelease t;
if ( !moreInitialSyncsPending && oplogReader.more() ) {
continue;
}
// otherwise, break out of loop so we can set to completed or clone more dbs
}
if( oplogReader.awaitCapable() && tailing )
okResultCode = 0; // don't sleep
syncedTo = nextOpTime;
save(); // note how far we are synced up to now
log() << "repl: applied " << n << " operations" << endl;
nApplied = n;
log() << "repl: end sync_pullOpLog syncedTo: " << syncedTo.toStringLong() << endl;
break;
}
else {
}
OCCASIONALLY if( n > 0 && ( n > 100000 || time(0) - saveLast > 60 ) ) {
// periodically note our progress, in case we are doing a lot of work and crash
dblock lk;
syncedTo = nextOpTime;
// can't update local log ts since there are pending operations from our peer
save();
log() << "repl: checkpoint applied " << n << " operations" << endl;
log() << "repl: syncedTo: " << syncedTo.toStringLong() << endl;
saveLast = time(0);
n = 0;
}
BSONObj op = oplogReader.next();
unsigned b = replApplyBatchSize;
bool justOne = b == 1;
scoped_ptr<writelock> lk( justOne ? 0 : new writelock() );
while( 1 ) {
BSONElement ts = op.getField("ts");
if( !( ts.type() == Date || ts.type() == Timestamp ) ) {
log() << "sync error: problem querying remote oplog record" << endl;
log() << "op: " << op.toString() << endl;
log() << "halting replication" << endl;
replInfo = replAllDead = "sync error: no ts found querying remote oplog record";
throw SyncException();
}
OpTime last = nextOpTime;
nextOpTime = OpTime( ts.date() );
if ( !( last < nextOpTime ) ) {
log() << "sync error: last applied optime at slave >= nextOpTime from master" << endl;
log() << " last: " << last.toStringLong() << endl;
log() << " nextOpTime: " << nextOpTime.toStringLong() << endl;
log() << " halting replication" << endl;
replInfo = replAllDead = "sync error last >= nextOpTime";
uassert( 10123 , "replication error last applied optime at slave >= nextOpTime from master", false);
}
if ( replSettings.slavedelay && ( unsigned( time( 0 ) ) < nextOpTime.getSecs() + replSettings.slavedelay ) ) {
assert( justOne );
oplogReader.putBack( op );
_sleepAdviceTime = nextOpTime.getSecs() + replSettings.slavedelay + 1;
dblock lk;
if ( n > 0 ) {
syncedTo = last;
save();
}
log() << "repl: applied " << n << " operations" << endl;
log() << "repl: syncedTo: " << syncedTo.toStringLong() << endl;
log() << "waiting until: " << _sleepAdviceTime << " to continue" << endl;
return okResultCode;
}
sync_pullOpLog_applyOperation(op, !justOne);
n++;
if( --b == 0 )
break;
// if to here, we are doing mulpile applications in a singel write lock acquisition
if( !oplogReader.moreInCurrentBatch() ) {
// break if no more in batch so we release lock while reading from the master
break;
}
op = oplogReader.next();
getDur().commitIfNeeded();
}
}
}
return okResultCode;
}
BSONObj userReplQuery = fromjson("{\"user\":\"repl\"}");
bool replAuthenticate(DBClientBase *conn) {
if( noauth ) {
return true;
}
if( ! cc().isAdmin() ) {
log() << "replauthenticate: requires admin permissions, failing\n";
return false;
}
string u;
string p;
if (internalSecurity.pwd.length() > 0) {
u = internalSecurity.user;
p = internalSecurity.pwd;
}
else {
BSONObj user;
{
dblock lk;
Client::Context ctxt("local.");
if( !Helpers::findOne("local.system.users", userReplQuery, user) ||
// try the first user in local
!Helpers::getSingleton("local.system.users", user) ) {
log() << "replauthenticate: no user in local.system.users to use for authentication\n";
return false;
}
}
u = user.getStringField("user");
p = user.getStringField("pwd");
massert( 10392 , "bad user object? [1]", !u.empty());
massert( 10393 , "bad user object? [2]", !p.empty());
}
string err;
if( !conn->auth("local", u.c_str(), p.c_str(), err, false) ) {
log() << "replauthenticate: can't authenticate to master server, user:" << u << endl;
return false;
}
return true;
}
bool replHandshake(DBClientConnection *conn) {
string myname = getHostName();
BSONObj me;
{
dblock l;
// local.me is an identifier for a server for getLastError w:2+
if ( ! Helpers::getSingleton( "local.me" , me ) ||
! me.hasField("host") ||
me["host"].String() != myname ) {
// clean out local.me
Helpers::emptyCollection("local.me");
// repopulate
BSONObjBuilder b;
b.appendOID( "_id" , 0 , true );
b.append( "host", myname );
me = b.obj();
Helpers::putSingleton( "local.me" , me );
}
}
BSONObjBuilder cmd;
cmd.appendAs( me["_id"] , "handshake" );
if (theReplSet) {
cmd.append("member", theReplSet->selfId());
}
BSONObj res;
bool ok = conn->runCommand( "admin" , cmd.obj() , res );
// ignoring for now on purpose for older versions
log(ok) << "replHandshake res not: " << ok << " res: " << res << endl;
return true;
}
bool OplogReader::commonConnect(const string& hostName) {
if( conn() == 0 ) {
_conn = shared_ptr<DBClientConnection>(new DBClientConnection( false, 0, 60*10 /* tcp timeout */));
string errmsg;
ReplInfo r("trying to connect to sync source");
if ( !_conn->connect(hostName.c_str(), errmsg) ||
(!noauth && !replAuthenticate(_conn.get())) ) {
resetConnection();
log() << "repl: " << errmsg << endl;
return false;
}
}
return true;
}
bool OplogReader::connect(string hostName) {
if (conn() != 0) {
return true;
}
if (commonConnect(hostName)) {
return replHandshake(_conn.get());
}
return false;
}
bool OplogReader::connect(const BSONObj& rid, const int from, const string& to) {
if (conn() != 0) {
return true;
}
if (commonConnect(to)) {
log() << "handshake between " << from << " and " << to << endl;
return passthroughHandshake(rid, from);
}
return false;
}
bool OplogReader::passthroughHandshake(const BSONObj& rid, const int f) {
BSONObjBuilder cmd;
cmd.appendAs( rid["_id"], "handshake" );
cmd.append( "member" , f );
BSONObj res;
return conn()->runCommand( "admin" , cmd.obj() , res );
}
/* note: not yet in mutex at this point.
returns >= 0 if ok. return -1 if you want to reconnect.
return value of zero indicates no sleep necessary before next call
*/
int ReplSource::sync(int& nApplied) {
_sleepAdviceTime = 0;
ReplInfo r("sync");
if ( !cmdLine.quiet ) {
Nullstream& l = log();
l << "repl: from ";
if( sourceName() != "main" ) {
l << "source:" << sourceName() << ' ';
}
l << "host:" << hostName << endl;
}
nClonedThisPass = 0;
// FIXME Handle cases where this db isn't on default port, or default port is spec'd in hostName.
if ( (string("localhost") == hostName || string("127.0.0.1") == hostName) && cmdLine.port == CmdLine::DefaultDBPort ) {
log() << "repl: can't sync from self (localhost). sources configuration may be wrong." << endl;
sleepsecs(5);
return -1;
}
if ( !oplogReader.connect(hostName) ) {
log(4) << "repl: can't connect to sync source" << endl;
return -1;
}
/*
// get current mtime at the server.
BSONObj o = conn->findOne("admin.$cmd", opTimeQuery);
BSONElement e = o.getField("optime");
if( e.eoo() ) {
log() << "repl: failed to get cur optime from master" << endl;
log() << " " << o.toString() << endl;
return false;
}
uassert( 10124 , e.type() == Date );
OpTime serverCurTime;
serverCurTime.asDate() = e.date();
*/
return sync_pullOpLog(nApplied);
}
/* --------------------------------------------------------------*/
/*
TODO:
_ source has autoptr to the cursor
_ reuse that cursor when we can
*/
/* returns: # of seconds to sleep before next pass
0 = no sleep recommended
1 = special sentinel indicating adaptive sleep recommended
*/
int _replMain(ReplSource::SourceVector& sources, int& nApplied) {
{
ReplInfo r("replMain load sources");
dblock lk;
ReplSource::loadAll(sources);
replSettings.fastsync = false; // only need this param for initial reset
}
if ( sources.empty() ) {
/* replication is not configured yet (for --slave) in local.sources. Poll for config it
every 20 seconds.
*/
log() << "no source given, add a master to local.sources to start replication" << endl;
return 20;
}
int sleepAdvice = 1;
for ( ReplSource::SourceVector::iterator i = sources.begin(); i != sources.end(); i++ ) {
ReplSource *s = i->get();
int res = -1;
try {
res = s->sync(nApplied);
bool moreToSync = s->haveMoreDbsToSync();
if( res < 0 ) {
sleepAdvice = 3;
}
else if( moreToSync ) {
sleepAdvice = 0;
}
else if ( s->sleepAdvice() ) {
sleepAdvice = s->sleepAdvice();
}
else
sleepAdvice = res;
}
catch ( const SyncException& ) {
log() << "caught SyncException" << endl;
return 10;
}
catch ( AssertionException& e ) {
if ( e.severe() ) {
log() << "replMain AssertionException " << e.what() << endl;
return 60;
}
else {
log() << "repl: AssertionException " << e.what() << '\n';
}
replInfo = "replMain caught AssertionException";
}
catch ( const DBException& e ) {
log() << "repl: DBException " << e.what() << endl;
replInfo = "replMain caught DBException";
}
catch ( const std::exception &e ) {
log() << "repl: std::exception " << e.what() << endl;
replInfo = "replMain caught std::exception";
}
catch ( ... ) {
log() << "unexpected exception during replication. replication will halt" << endl;
replAllDead = "caught unexpected exception during replication";
}
if ( res < 0 )
s->oplogReader.resetConnection();
}
return sleepAdvice;
}
void replMain() {
ReplSource::SourceVector sources;
while ( 1 ) {
int s = 0;
{
dblock lk;
if ( replAllDead ) {
// throttledForceResyncDead can throw
if ( !replSettings.autoresync || !ReplSource::throttledForceResyncDead( "auto" ) ) {
log() << "all sources dead: " << replAllDead << ", sleeping for 5 seconds" << endl;
break;
}
}
assert( syncing == 0 ); // i.e., there is only one sync thread running. we will want to change/fix this.
syncing++;
}
try {
int nApplied = 0;
s = _replMain(sources, nApplied);
if( s == 1 ) {
if( nApplied == 0 ) s = 2;
else if( nApplied > 100 ) {
// sleep very little - just enought that we aren't truly hammering master
sleepmillis(75);
s = 0;
}
}
}
catch (...) {
out() << "caught exception in _replMain" << endl;
s = 4;
}
{
dblock lk;
assert( syncing == 1 );
syncing--;
}
if( relinquishSyncingSome ) {
relinquishSyncingSome = 0;
s = 1; // sleep before going back in to syncing=1
}
if ( s ) {
stringstream ss;
ss << "repl: sleep " << s << " sec before next pass";
string msg = ss.str();
if ( ! cmdLine.quiet )
log() << msg << endl;
ReplInfo r(msg.c_str());
sleepsecs(s);
}
}
}
static void replMasterThread() {
sleepsecs(4);
Client::initThread("replmaster");
int toSleep = 10;
while( 1 ) {
sleepsecs( toSleep );
/* write a keep-alive like entry to the log. this will make things like
printReplicationStatus() and printSlaveReplicationStatus() stay up-to-date
even when things are idle.
*/
{
writelocktry lk("",1);
if ( lk.got() ) {
toSleep = 10;
replLocalAuth();
try {
logKeepalive();
}
catch(...) {
log() << "caught exception in replMasterThread()" << endl;
}
}
else {
log(5) << "couldn't logKeepalive" << endl;
toSleep = 1;
}
}
}
}
void replSlaveThread() {
sleepsecs(1);
Client::initThread("replslave");
cc().iAmSyncThread();
{
dblock lk;
replLocalAuth();
}
while ( 1 ) {
try {
replMain();
sleepsecs(5);
}
catch ( AssertionException& ) {
ReplInfo r("Assertion in replSlaveThread(): sleeping 5 minutes before retry");
problem() << "Assertion in replSlaveThread(): sleeping 5 minutes before retry" << endl;
sleepsecs(300);
}
catch ( DBException& e ) {
problem() << "exception in replSlaveThread(): " << e.what()
<< ", sleeping 5 minutes before retry" << endl;
sleepsecs(300);
}
catch ( ... ) {
problem() << "error in replSlaveThread(): sleeping 5 minutes before retry" << endl;
sleepsecs(300);
}
}
}
void tempThread() {
while ( 1 ) {
out() << dbMutex.info().isLocked() << endl;
sleepmillis(100);
}
}
void newRepl();
void oldRepl();
void startReplSets(ReplSetCmdline*);
void startReplication() {
/* if we are going to be a replica set, we aren't doing other forms of replication. */
if( !cmdLine._replSet.empty() ) {
if( replSettings.slave || replSettings.master ) {
log() << "***" << endl;
log() << "ERROR: can't use --slave or --master replication options with --replSet" << endl;
log() << "***" << endl;
}
newRepl();
replSet = true;
ReplSetCmdline *replSetCmdline = new ReplSetCmdline(cmdLine._replSet);
boost::thread t( boost::bind( &startReplSets, replSetCmdline) );
return;
}
oldRepl();
/* this was just to see if anything locks for longer than it should -- we need to be careful
not to be locked when trying to connect() or query() the other side.
*/
//boost::thread tempt(tempThread);
if( !replSettings.slave && !replSettings.master )
return;
{
dblock lk;
replLocalAuth();
}
if ( replSettings.slave ) {
assert( replSettings.slave == SimpleSlave );
log(1) << "slave=true" << endl;
boost::thread repl_thread(replSlaveThread);
}
if ( replSettings.master ) {
log(1) << "master=true" << endl;
replSettings.master = true;
createOplog();
boost::thread t(replMasterThread);
}
while( replSettings.fastsync ) // don't allow writes until we've set up from log
sleepmillis( 50 );
}
void testPretouch() {
int nthr = min(8, 8);
nthr = max(nthr, 1);
int m = 8 / nthr;
ThreadPool tp(nthr);
vector<BSONObj> v;
BSONObj x = BSON( "ns" << "test.foo" << "o" << BSON( "_id" << 1 ) << "op" << "i" );
v.push_back(x);
v.push_back(x);
v.push_back(x);
unsigned a = 0;
while( 1 ) {
if( a >= v.size() ) break;
unsigned b = a + m - 1; // v[a..b]
if( b >= v.size() ) b = v.size() - 1;
tp.schedule(pretouchN, v, a, b);
DEV cout << "pretouch task: " << a << ".." << b << endl;
a += m;
}
tp.join();
}
class ReplApplyBatchSizeValidator : public ParameterValidator {
public:
ReplApplyBatchSizeValidator() : ParameterValidator( "replApplyBatchSize" ) {}
virtual bool isValid( BSONElement e , string& errmsg ) {
int b = e.numberInt();
if( b < 1 || b > 1024 ) {
errmsg = "replApplyBatchSize has to be >= 1 and < 1024";
return false;
}
if ( replSettings.slavedelay != 0 && b > 1 ) {
errmsg = "can't use a batch size > 1 with slavedelay";
return false;
}
if ( ! replSettings.slave ) {
errmsg = "can't set replApplyBatchSize on a non-slave machine";
return false;
}
return true;
}
} replApplyBatchSizeValidator;
} // namespace mongo
|