cluster.c
48 KB
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
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
/*
clsync - file tree sync utility based on inotify
Copyright (C) 2013 Dmitry Yu Okunev <dyokunev@ut.mephi.ru> 0x8E30679C
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
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 for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
/*
Hello, dear developer.
You can ask me directly by e-mail or IRC, if something seems too
hard.
-- 0x8E30679C
*/
#ifdef CLUSTER_SUPPORT
#include "common.h"
#include "indexes.h"
#include "error.h"
#include "cluster.h"
#include "sync.h"
#include "calc.h"
#include "malloc.h"
// Global variables. They will be initialized in cluster_init()
#define CLUSTER_RECV_PROC_ERRLIMIT (1<<8)
#define NODES_ALLOC (MAX(MAXNODES, NODEID_NOID)+1)
int sock_i = -1;
struct sockaddr_in sa_i = {0};
int sock_o = -1;
struct sockaddr_in sa_o = {0};
ctx_t *ctx_p = NULL;
indexes_t *indexes_p = NULL;
pthread_t pthread_cluster = 0;
nodeinfo_t nodeinfo[NODES_ALLOC]= {{0}};
nodeinfo_t *nodeinfo_my = NULL;
uint8_t node_id_my = NODEID_NOID;
uint8_t node_ids[NODES_ALLOC] = {0};
unsigned int cluster_timeout = 0;
uint8_t node_count = 0;
uint8_t node_online = 0;
cluster_recvproc_funct_t recvproc_funct[COUNT_CLUSTERCMDID] = {NULL};
window_t window_i = {0};
window_t window_o = {0};
/**
* @brief Adds command (message) to window_p->buffer
*
* @param[in] clustercmd_p Pointer to cluster cmd to put into window
*
* @retval zero Successful
* @retval non-zero If got error while deleting the message. The error-code is placed into returned value.
*
*/
static inline int clustercmd_window_add(window_t *window_p, clustercmd_t *clustercmd_p, GHashTable *serial2queuedpacket_ht) {
#ifdef PARANOID
if (clustercmd_p->h.src_node_id >= MAXNODES) {
error("Invalid src_node_id: %i.", clustercmd_p->h.src_node_id);
return EINVAL;
}
#endif
// Checking if there enough window_p->cells allocated
if (window_p->packets_len >= window_p->size) {
window_p->size += ALLOC_PORTION;
# define CXREALLOC(a, size) \
(typeof(a))xrealloc((char *)(a), (size_t)(size) * sizeof(*(a)))
window_p->packets_id = CXREALLOC(window_p->packets_id, window_p->size);
window_p->occupied_sides = CXREALLOC(window_p->occupied_sides, window_p->size);
# undef CXREALLOC
}
// Calculating required memory space in buffer for the message
size_t clustercmd_size = CLUSTERCMD_SIZE(clustercmd_p);
size_t required_space = sizeof(clustercmdqueuedpackethdr_t) + clustercmd_size;
// Searching occupied boundaries in the window_p->buffer
size_t occupied_left = SIZE_MAX, occupied_right=0;
int i;
i = 0;
while (i < window_p->packets_len) {
unsigned int window_id;
window_id = window_p->packets_id[ i++ ];
occupied_left = MIN(occupied_left, window_p->occupied_sides[window_id].left);
occupied_right = MAX(occupied_right, window_p->occupied_sides[window_id].right);
}
debug(3, "w.size == %u, b_left == %u; b_right == %u; w.buf_size == %u; r_space == %u",
window_p->size, occupied_left, occupied_right, window_p->buf_size, required_space);
// Trying to find a space in the buffer to place message
size_t buf_coordinate = SIZE_MAX;
if (window_p->packets_len) {
// Free space from left (start of buffer)
size_t free_left = occupied_left;
// Free space from right (end of buffer)
size_t free_right = window_p->buf_size - occupied_right;
if (free_left > required_space)
buf_coordinate = free_left - required_space;
else
if (free_right > required_space)
buf_coordinate = occupied_right;
else
{
// Not enough space in the window_p->buffer;
window_p->buf_size += MAX(CLUSTER_WINDOW_BUFSIZE_PORTION, required_space);
window_p->buf = xrealloc(window_p->buf, window_p->buf_size);
buf_coordinate = occupied_right;
}
debug(3, "f_left == %u; f_right == %u; b_coord == %u; w.buf_size == %u",
free_left, free_right, buf_coordinate, window_p->buf_size);
} else {
buf_coordinate = 0;
if (window_p->buf_size <= required_space) {
window_p->buf_size += MAX(CLUSTER_WINDOW_BUFSIZE_PORTION, required_space);
window_p->buf = xrealloc(window_p->buf, window_p->buf_size);
}
}
unsigned int window_id;
// packet id in window
window_id = window_p->packets_len;
// reserving the space in buffer
window_p->occupied_sides[window_id].left = buf_coordinate;
window_p->occupied_sides[window_id].right = buf_coordinate + required_space;
// placing information into buffer
clustercmdqueuedpacket_t *queuedpacket_p;
debug(3, "b_coord == %u", buf_coordinate);
queuedpacket_p = (clustercmdqueuedpacket_t *)&window_p->buf[buf_coordinate];
memset(&queuedpacket_p->h, 0, sizeof(queuedpacket_p->h));
memcpy(&queuedpacket_p->cmd, clustercmd_p, clustercmd_size);
queuedpacket_p->h.window_id = window_id;
// remembering new packet
g_hash_table_insert(serial2queuedpacket_ht, GINT_TO_POINTER(clustercmd_p->h.serial), queuedpacket_p);
window_p->packets_id[window_p->packets_len++] = window_id;
return 0;
}
/**
* @brief Removes command (message) from window_p->buffer
*
* @param[in] queuedpacket_p Pointer to queuedpacket structure of the command (message)
*
* @retval zero Successful
* @retval non-zero If got error while deleting the message. The error-code is placed into returned value.
*
*/
static inline int clustercmd_window_del(window_t *window_p, clustercmdqueuedpacket_t *queuedpacket_p, GHashTable *serial2queuedpacket_ht) {
#ifdef PARANOID
if(!window_p->size) {
error("window not allocated.");
return EINVAL;
}
if(!window_p->packets_len) {
error("there already no packets in the window.");
return EINVAL;
}
#endif
unsigned int window_id_del = queuedpacket_p->h.window_id;
unsigned int window_id_last = --window_p->packets_len;
// Forgeting the packet
// Moving the last packet into place of deleting packet, to free the tail in "window_p->packets_id" and "window_p->occupied_sides"
if(window_id_del != window_id_last) {
debug(3, "%i -> %i", window_id_last, window_id_del);
window_p->packets_id[window_id_del] = window_p->packets_id[window_id_last];
memcpy(&window_p->occupied_sides[window_id_del], &window_p->occupied_sides[window_id_last], sizeof(window_p->occupied_sides[window_id_del]));
}
// Removing from hash table
g_hash_table_remove(serial2queuedpacket_ht, GINT_TO_POINTER(queuedpacket_p->cmd.h.serial));
return 0;
}
/**
* @brief Calculates Adler32 for clustercmd
*
* @param[in] clustercmd_p Pointer to clustercmd
* @param[out] clustercmdadler32_p Pointer to structure to return value(s)
*
* @retval zero On successful calculation
* @retval non-zero On error. Error-code is placed into returned value.
*
*/
int clustercmd_adler32_calc(clustercmd_t *clustercmd_p, clustercmdadler32_t *clustercmdadler32_p, adler32_calc_t flags) {
debug(15, "(%p, %p, 0x%x)", clustercmd_p, clustercmdadler32_p, flags);
if (flags & ADLER32_CALC_DATA) {
uint32_t adler32;
uint32_t size = clustercmd_p->h.data_len;
char *ptr = clustercmd_p->data.p;
// Calculating
adler32 = adler32_calc((unsigned char *)ptr, CLUSTER_PAD(size));
debug(20, "dat: 0x%x", adler32);
// Ending
clustercmdadler32_p->dat = adler32 ^ 0xFFFFFFFF;
}
if (flags & ADLER32_CALC_HEADER) {
uint32_t adler32;
clustercmdadler32_t adler32_save;
// Preparing
memcpy(&adler32_save.hdr, &clustercmd_p->h.adler32.hdr, sizeof(clustercmd_p->h.adler32.hdr));
memset(&clustercmd_p->h.adler32.hdr, 0, sizeof(clustercmd_p->h.adler32.hdr));
adler32 = 0xFFFFFFFF;
uint32_t size = sizeof(clustercmdhdr_t);
char *ptr = (char *)&clustercmd_p->h;
// Calculating
adler32 = adler32_calc((unsigned char *)ptr, size);
debug(20, "hdr: 0x%x", adler32);
// Ending
memcpy(&clustercmd_p->h.adler32.hdr, &adler32_save.hdr, sizeof(clustercmd_p->h.adler32.hdr));
clustercmdadler32_p->hdr = adler32 ^ 0xFFFFFFFF;
}
return 0;
}
/**
* @brief Sends message to another nodes of the cluster.
*
* @param[in] clustercmd_p Command structure pointer.
*
* @retval zero Successfully send.
* @retval non-zero Got error, while sending.
*
*/
int _cluster_send(clustercmd_t *clustercmd_p) {
debug(10, "{h.dst_node_id: %u, h.src_node_id: %u, cmd_id: %u, data_len: %u}",
clustercmd_p->h.dst_node_id, clustercmd_p->h.src_node_id, clustercmd_p->h.cmd_id,
clustercmd_p->h.data_len);
clustercmd_p->h.src_node_id = node_id_my;
SAFE (clustercmd_adler32_calc(clustercmd_p, &clustercmd_p->h.adler32, ADLER32_CALC_ALL), return _SAFE_rc );
debug(3, "Sending: "
"{h.dst_node_id: %u, h.src_node_id: %u, cmd_id: %u, adler32.hdr: %p, adler32.dat: %p, data_len: %u}",
clustercmd_p->h.dst_node_id, clustercmd_p->h.src_node_id, clustercmd_p->h.cmd_id,
(void *)(long)clustercmd_p->h.adler32.hdr, (void *)(long)clustercmd_p->h.adler32.dat,
clustercmd_p->h.data_len);
nodeinfo_t *nodeinfo_p;
nodeinfo_p = &nodeinfo[clustercmd_p->h.dst_node_id];
// Checking if the node online
switch (nodeinfo_p->status) {
case NODESTATUS_DOESNTEXIST:
case NODESTATUS_OFFLINE:
debug(1, "There's no online node with id %u. Skipping sending.", clustercmd_p->h.dst_node_id);
return EADDRNOTAVAIL;
default:
break;
}
// Putting the message into an output window
if (nodeinfo_my != NULL)
clustercmd_window_add(&window_o, clustercmd_p, nodeinfo_my->serial2queuedpacket_ht);
// Sending the message
debug(10, "sendto(%p, %p, %i, 0, %p, %i)", sock_o, clustercmd_p, CLUSTERCMD_SIZE_PADDED(clustercmd_p), &sa_o, sizeof(sa_o));
debug(50, "clustercmd_p->data.p[0] == 0x%x", clustercmd_p->data.p[0]);
critical_on (sendto(sock_o, clustercmd_p, CLUSTERCMD_SIZE_PADDED(clustercmd_p), 0, &sa_o, sizeof(sa_o)) == -1);
// Finishing
return 0;
}
static inline int cluster_send(clustercmd_t *clustercmd_p) {
int rc;
rc = _cluster_send(clustercmd_p);
debug(10, "__________________sent___________________ %i", rc);
return rc;
}
/**
* @brief Changes information about node's status in nodeinfo[] and updates connected information.
*
* @param[in] node_id node_id of the node.
* @param[in] node_status New node status.
*
* @retval zero Successful
* @retval non-zero If got error while changing the status. The error-code is placed into returned value.
*
*/
int node_update_status(uint8_t node_id, uint8_t node_status) {
debug(3, "%i, %i", node_id, node_status);
uint8_t node_status_old = nodeinfo[node_id].status;
nodeinfo_t *nodeinfo_p = &nodeinfo[node_id];
if ((node_status == NODESTATUS_DOESNTEXIST) && (node_status_old != NODESTATUS_DOESNTEXIST)) {
node_count--;
node_ids[nodeinfo_p->num] = node_ids[node_count];
g_hash_table_destroy(nodeinfo_p->modtime_ht);
g_hash_table_destroy(nodeinfo_p->serial2queuedpacket_ht);
#ifdef VERYPARANOID
memset(nodeinfo_p, 0, sizeof(*nodeinfo_p));
#endif
return 0;
}
if (node_status == node_status_old)
return 0;
switch (node_status_old) {
case NODESTATUS_DOESNTEXIST:
nodeinfo_p->id = node_id;
nodeinfo_p->num = node_count;
nodeinfo_p->last_serial = -1;
nodeinfo_p->modtime_ht = g_hash_table_new_full(g_str_hash, g_str_equal, free, 0);
nodeinfo_p->serial2queuedpacket_ht = g_hash_table_new_full(g_direct_hash, g_direct_equal, 0, 0);
node_ids[node_count] = node_id;
node_count++;
#ifdef PARANOID
if (node_status == NODESTATUS_OFFLINE)
break; // In case of NODESTATUS_DOESNTEXIST -> NODESTATUS_OFFLINE, node_online should be increased
#endif
case NODESTATUS_OFFLINE:
nodeinfo_p->last_serial = -1;
node_online++;
break;
default:
if (node_status == NODESTATUS_OFFLINE)
node_online--;
break;
}
nodeinfo[node_id].status = node_status;
return 0;
}
/**
* @brief Changes information about node's name and updates connected information.
*
* @param[in] node_id node_id of the node.
* @param[in] node_name New node name
*
* @retval zero Successful
* @retval non-zero If got error while changing the status. The error-code is placed into returned value.
*
*/
int node_update_name(uint8_t node_id, const char *node_name) {
debug(3, "%i, \"%s\"", node_id, node_name);
nodeinfo_t *nodeinfo_p = &nodeinfo[node_id];
void *ret = g_hash_table_lookup(indexes_p->nodenames_ht, node_name);
if (ret != NULL) {
int rc;
uint8_t node_old_id = GPOINTER_TO_INT(ret);
nodeinfo_t *nodeinfo_old_p = &nodeinfo[node_old_id];
debug(5, "nodename \"%s\" had been used by node_id == %i", node_name, node_old_id);
if (node_old_id == node_id) {
debug(10, "node_old_id [%i] == node_id [%i]", node_old_id, node_id);
return 0;
}
{
debug(15, "Sending a DIE command to node_id (%i)", node_old_id);
clustercmd_t *clustercmd_p = CLUSTER_ALLOCA(clustercmd_die_t, 0);
clustercmd_p->h.cmd_id = CLUSTERCMDID_DIE;
clustercmd_p->h.dst_node_id = node_old_id;
if ((rc = cluster_send(clustercmd_p)))
return rc;
}
nodeinfo_old_p->node_name = NULL;
debug(3, "changing status of node_id == %i to NODESTATUS_OFFLINE", node_old_id);
node_update_status(node_old_id, NODESTATUS_OFFLINE);
}
char *node_name_dup = strdup(node_name);
nodeinfo_p->node_name = node_name_dup;
g_hash_table_replace(indexes_p->nodenames_ht, node_name_dup, GINT_TO_POINTER(node_id));
return 0;
}
/**
* @brief Sets message processing functions for cluster_recv_proc() function for specified command type
*
* @param[in] cmd_id The command type
* @param[in] procfunct The processing function for messages with specified cmd_id
*
* @retval zero Successful
* @retval non-zero If got error while setting processing function. The error-code is placed into returned value.
*
*/
static inline int cluster_recv_proc_set(clustercmd_id_t cmd_id, cluster_recvproc_funct_t procfunct) {
recvproc_funct[cmd_id] = procfunct;
return 0;
}
/**
* @brief Safe wrapper for recvfrom() function
*
* @param[in] sock The socket descriptor
* @param[out] _buf_p Pointer to pointer to buffer
* @param[in] flags Flags...
*
* @retval zero Successful
* @retval non-zero If got error while read()-ing. The error-code is placed into returned value. "-1" means that message is too short.
*
*/
static inline int cluster_read(int sock, clustercmd_t **cmd_pp, cluster_read_flags_t flags) {
static char buf[CLUSTER_PACKET_MAXSIZE];
static clustercmd_t *cmd_p = (void *)buf;
struct sockaddr_in sa_in;
size_t sa_in_len = sizeof(sa_in);
*cmd_pp = cmd_p;
debug(20, "%i, %p, 0x%x", sock, buf, flags);
int readret = recvfrom(sock, buf, CLUSTER_PACKET_MAXSIZE, MSG_WAITALL, (struct sockaddr *)&sa_in, (socklen_t * restrict)&sa_in_len);
debug(30, "recvfrom(%i, %p, 0x%x, %p, %p) -> %i", sock, buf, MSG_WAITALL, &sa_in, &sa_in_len, readret);
#ifdef PARANOID
if (!readret) {
error("recvfrom() returned 0. This shouldn't happend. Exit.");
return EINVAL;
}
#endif
if (readret < 0) {
error("recvfrom() returned %i. "
"Seems, that something wrong with network socket.",
readret);
return errno != -1 ? errno : -2;
}
debug(2, "Got message from %s (len: %i).", inet_ntoa(sa_in.sin_addr), readret);
if (readret < sizeof(clustercmdhdr_t)) {
// Too short message
error("Warning: cluster_read(): Got too short message from node (no header [or too short]). Ignoring it.");
return -1;
}
if (readret < CLUSTERCMD_SIZE(cmd_p)) {
// Too short message
error("Warning: cluster_read(): Got too short message from node (no data [or too short]). Ignoring it.");
return -1;
}
// Incorrect size?
if (readret & 0x3) {
error("Warning: cluster_recv(): Received packet of size %i (not a multiple of 4). Ignoring it.", readret);
return 0;
}
return 0;
}
/**
* @brief Sends packet-reject notification
*
* @param[in] clustercmd_p Pointer to clustercmd that will be rejected
* @param[in] reason Reason why the clustercmd is denied
*
* @retval zero Successful
* @retval non-zero If got error while read()-ing. The error-code is placed into returned value. "-1" means that message is too short.
*
*/
static inline int clustercmd_reject(clustercmd_t *clustercmd_p, uint8_t reason) {
clustercmd_t *clustercmd_rej_p = CLUSTER_ALLOCA(clustercmd_rej_t, 0);
clustercmd_rej_p->h.dst_node_id = clustercmd_p->h.src_node_id;
clustercmd_rej_p->data.rej.serial = clustercmd_p->h.serial;
clustercmd_rej_p->data.rej.reason = reason;
return cluster_send(clustercmd_rej_p);
}
#define CLUSTER_RECV_RETURNMESSAGE(clustercmd_p) {\
last_serial = (clustercmd_p)->h.serial;\
last_src_node_id = (clustercmd_p)->h.src_node_id;\
if(clustercmd_pp != NULL)\
*clustercmd_pp = (clustercmd_p);\
return 1;\
}
/**
* @brief Receives message from another nodes of the cluster. (not thread-safe)
*
* @param[out] clustercmd_pp Pointer to command structure pointer. It will be re-allocated every time when size is not enough. Allocated space will be reused on next calling.
* @param[i] timeout Timeout (in milliseconds).
*
* @retval 1 If there's new message.
* @retval 0 If there's no new messages.
* @retval -1 If got error while receiving. The error-code is placed into "errno".
*
*/
static int _cluster_recv(clustercmd_t **clustercmd_pp, struct timeval *timeout) {
clustercmd_t *clustercmd_p;
static uint8_t last_src_node_id = NODEID_NOID;
static uint32_t last_serial = 0;
// Checking if there message is waiting in the window
if (last_src_node_id != NODEID_NOID) {
nodeinfo_t *nodeinfo_p = &nodeinfo[last_src_node_id];
if (nodeinfo_p->serial2queuedpacket_ht != NULL) {
clustercmdqueuedpacket_t *clustercmdqueuedpacket_p = (clustercmdqueuedpacket_t *)
g_hash_table_lookup(nodeinfo_p->serial2queuedpacket_ht, GINT_TO_POINTER(last_serial+1));
if (clustercmdqueuedpacket_p != NULL)
CLUSTER_RECV_RETURNMESSAGE(&clustercmdqueuedpacket_p->cmd);
}
}
// Checking if there any event on read socket
// select()
fd_set rfds;
FD_ZERO(&rfds);
FD_SET(sock_i, &rfds);
debug(3, "select() with timeout {%u, %u}", timeout->tv_sec, timeout->tv_usec);
int selret = select(sock_i+1, &rfds, NULL, NULL, timeout);
// processing select()'s retuned value
if (selret < 0) {
error("got error while select().");
return -1;
}
if (selret == 0) {
debug(3, "no new messages.");
return 0;
}
debug(3, "got new message(s).");
debug(10, "Reading new message's header");
clustercmdadler32_t adler32;
//clustercmd_t *clustercmd_p = (clustercmd_t *)mmap(NULL, sizeof(clustercmdhdr_t), PROT_NONE,
// MAP_PRIVATE, sock, 0);
int ret;
if ((ret=cluster_read(sock_i, clustercmd_pp, CLREAD_NONE))) {
if(ret == -1) return 0; // Invalid message? Skipping.
error("Got error from cluster_read().");
errno = ret;
return -1;
}
clustercmd_p = *clustercmd_pp;
debug(3, "Received: {h.dst_node_id: %u, h.src_node_id: %u, cmd_id: %u,"
" serial: %u, adler32: {0x%x, 0x%x}, data_len: %u}",
clustercmd_p->h.dst_node_id, clustercmd_p->h.src_node_id,
clustercmd_p->h.cmd_id, clustercmd_p->h.serial,
clustercmd_p->h.adler32.hdr, clustercmd_p->h.adler32.dat,
clustercmd_p->h.data_len);
// Checking adler32 of packet headers.
clustercmd_adler32_calc(clustercmd_p, &adler32, ADLER32_CALC_HEADER);
if (adler32.hdr != clustercmd_p->h.adler32.hdr) {
debug(1, "hdr-adler32 mismatch: %p != %p.",
(void*)(long)clustercmd_p->h.adler32.hdr, (void*)(long)adler32.hdr);
if((ret=clustercmd_reject(clustercmd_p, REJ_ADLER32MISMATCH)) != EADDRNOTAVAIL) {
error("Got error while clustercmd_reject().");
errno = ret;
return -1;
}
}
// Checking src_node_id and dst_node_id
uint8_t src_node_id = clustercmd_p->h.src_node_id;
uint8_t dst_node_id = clustercmd_p->h.dst_node_id;
// Packet from registering node?
if (src_node_id == NODEID_NOID) {
// Wrong command from registering node?
if (clustercmd_p->h.cmd_id != CLUSTERCMDID_HELLO) {
error("Warning: cluster_recv(): Got non hello packet from NOID node. Ignoring the packet.");
return 0;
}
if (clustercmd_p->h.serial != 0) {
error("Warning: cluster_recv(): Got packet with non-zero serial from NOID node. Ignoring the packet.");
return 0;
}
} else
// Wrong src_node_id?
if (src_node_id >= MAXNODES) {
error("Warning: cluster_recv(): Invalid h.src_node_id: %i >= "XTOSTR(MAXNODES)"",
src_node_id);
return -1;
}
// Is this broadcast message?
if (dst_node_id == NODEID_NOID) {
// CODE HERE
} else
// Wrong dst_node_id?
if (dst_node_id >= MAXNODES) {
error("Warning: cluster_recv(): Invalid h.dst_node_id: %i >= "XTOSTR(MAXNODES)"",
dst_node_id);
return -1;
}
// Seems, that headers are correct. Continuing.
// Paranoid routines
// The message from us? Something wrong if it is.
if ((clustercmd_p->h.src_node_id == node_id_my) && (node_id_my != NODEID_NOID))
critical("node_id collision");
nodeinfo_t *nodeinfo_p = &nodeinfo[src_node_id];
// Not actual packet?
long serial_diff = clustercmd_p->h.serial - nodeinfo_p->last_serial;
debug(10, "serial_diff == %i", serial_diff);
if (serial_diff <= 0 || serial_diff > CLUSTER_WINDOW_PCKTLIMIT) {
debug(1, "Ignoring packet (serial %i) from %i due to serial_diff: 0 <= || > %i",
clustercmd_p->h.serial, src_node_id, serial_diff, CLUSTER_WINDOW_PCKTLIMIT);
return -1;
}
// Is this misordered packet?
if (clustercmd_p->h.serial != nodeinfo_p->last_serial + 1) {
clustercmd_window_add(&window_i, clustercmd_p, nodeinfo_p->serial2queuedpacket_ht);
return -1;
}
// Is this the end of packet (packet without data)
if (clustercmd_p->h.data_len == 0)
CLUSTER_RECV_RETURNMESSAGE(clustercmd_p);
// Too big data?
if (clustercmd_p->h.data_len > CLUSTER_PACKET_MAXSIZE) {
error("Warning: cluster_recv(): Got too big message from node %i. Ignoring it.",
src_node_id);
return -1;
}
/*
// Need more space for this packet?
if (CLUSTERCMD_SIZE(clustercmd_p) > size) {
size = CLUSTERCMD_SIZE(clustercmd_p);
clustercmd_p = (clustercmd_t *)xrealloc((char *)clustercmd_p, size);
}
debug(10, "Reading the data");
if ((ret=cluster_read(sock_i, (void *)clustercmd_p->data.p, CLUSTER_PAD(clustercmd_p->h.data_len), CLREAD_CONTINUE))) {
if (ret == -1) return 0;
error("Got error from cluster_read().");
errno = ret;
return -1;
}
*/
// Checking adler32 of packet data.
clustercmd_adler32_calc(clustercmd_p, &adler32, ADLER32_CALC_DATA);
if (adler32.dat != clustercmd_p->h.adler32.dat) {
debug(1, "dat-adler32 mismatch: %p != %p.",
(void*)(long)clustercmd_p->h.adler32.dat, (void*)(long)adler32.dat);
if ((ret=clustercmd_reject(clustercmd_p, REJ_ADLER32MISMATCH)) != EADDRNOTAVAIL) {
error("Got error while clustercmd_reject().");
errno = ret;
return -1;
}
}
CLUSTER_RECV_RETURNMESSAGE(clustercmd_p);
}
static inline int cluster_recv(clustercmd_t **clustercmd_pp, struct timeval *timeout) {
int rc;
rc = _cluster_recv(clustercmd_pp, timeout);
debug(10, "__________________recv___________________ %i", rc);
return rc;
}
/**
* @brief (hsyncop) Reads messages for time "_timeout" and proceeding them to recvproc_funct[] functions
*
* @param[in] _timeout How long to wait messages (totally)
*
* @retval zero Successful
* @retval non-zero If got error while reading or processing messages. The error-code is placed into returned value.
*
*/
int cluster_recv_proc(unsigned int timeout_rel_int) {
struct timeval timeout_rel, timeout_abs, tv_abs;
int error_count, iteration;
debug(3, "cluster_recv_proc(%i)", timeout_rel_int);
clustercmd_t *clustercmd_p = NULL;
int ret;
timeout_rel.tv_sec = timeout_rel_int / 1000;
timeout_rel.tv_usec = timeout_rel_int % 1000;
gettimeofday(&tv_abs, NULL);
timeradd(&tv_abs, &timeout_rel, &timeout_abs);
error_count = 0;
iteration = 0;
while (1) {
if (iteration++) {
gettimeofday(&tv_abs, NULL);
if (timercmp(&timeout_abs, &tv_abs, >))
timersub(&timeout_abs, &tv_abs, &timeout_rel);
else
memset(&timeout_rel, 0, sizeof(timeout_rel));
debug(5, "timeout_abs == {%u, %u}; tv_abs == {%u, %u}; timeout_rel == {%u, %u}",
timeout_abs.tv_sec, timeout_abs.tv_usec,
tv_abs.tv_sec, tv_abs.tv_usec,
timeout_rel.tv_sec, timeout_rel.tv_usec
);
}
ret = cluster_recv(&clustercmd_p, &timeout_rel);
if (!ret)
break;
// Exit if error
if (ret == -1) {
warning("Got error while cluster_recv(). error_count == %i", error_count);
error_count++;
critical_on (error_count >= CLUSTER_RECV_PROC_ERRLIMIT);
}
// If we have appropriate callback function, then call it! :)
if (recvproc_funct[clustercmd_p->h.cmd_id] != NULL) {
debug(2, "Calling function by pointer %p", recvproc_funct[clustercmd_p->h.cmd_id]);
if ((ret = recvproc_funct[clustercmd_p->h.cmd_id](clustercmd_p))) {
error("Got error from recvproc_funct[%i]: %s (%i)",
clustercmd_p->h.cmd_id);
return ret;
}
continue;
}
// We didn't found an appropriate callback function
debug(2, "There's no appropriate callback function for cmd_id == %i", clustercmd_p->h.cmd_id);
}
return 0;
}
/**
* @brief recvproc-function for DIE-messages
*
* @param[in] clustercmd_p Pointer to clustercmd
*
* @retval zero Successfully initialized. (never happens)
* @retval non-zero Got error, while initializing.
*
*/
static int cluster_recvproc_die(clustercmd_t *clustercmd_p) {
critical("Got DIE message from node_id == %i,", clustercmd_p->h.src_node_id);
return 0;
}
/**
* @brief recvproc-function for ACK-messages
*
* @param[in] clustercmd_p Pointer to clustercmd
*
* @retval zero Successfully initialized.
* @retval non-zero Got error, while initializing.
*
*/
static int cluster_recvproc_ack(clustercmd_t *clustercmd_p) {
uint32_t cmd_serial_ack = clustercmd_p->data.ack.serial;
clustercmdqueuedpacket_t *queuedpacket_p =
(clustercmdqueuedpacket_t *)g_hash_table_lookup(nodeinfo_my->serial2queuedpacket_ht, GINT_TO_POINTER(cmd_serial_ack));
if (queuedpacket_p == NULL)
return 0;
uint8_t node_id_from = clustercmd_p->h.src_node_id;
if (! queuedpacket_p->h.w.o.ack_from[node_id_from]) {
queuedpacket_p->h.w.o.ack_count++;
queuedpacket_p->h.w.o.ack_from[node_id_from]++;
if (queuedpacket_p->h.w.o.ack_count == node_count-1)
clustercmd_window_del(&window_o, queuedpacket_p, nodeinfo_my->serial2queuedpacket_ht);
}
return 0;
}
/**
* @brief Sets message processing functions for cluster_recv_proc() function for specified command type
*
* @param[in] cmd_id The command type
* @param[in] procfunct The processing function for messages with specified cmd_id
*
* @retval zero Successful
* @retval non-zero If got error while setting processing function. The error-code is placed into returned value.
*
*/
int cluster_io_init() {
cluster_recv_proc_set(CLUSTERCMDID_ACK, cluster_recvproc_ack);
cluster_recv_proc_set(CLUSTERCMDID_DIE, cluster_recvproc_die);
return 0;
}
/**
* @brief Antagonist of cluster_recv_proc_init() function. Freeing everything what was allocated in cluster_recv_proc_init()
*
* @retval zero Successfully initialized
* @retval non-zero Got error, while initializing
*
*/
int cluster_io_deinit() {
if(window_i.buf_size) {
#ifdef PARANOID
if(window_i.buf == NULL) {
error("window_i.buf_size != 0, but window_i.buf == NULL.");
} else
#endif
free(window_i.buf);
}
if(window_o.buf_size) {
#ifdef PARANOID
if(window_o.buf == NULL) {
error("window_o.buf_size != 0, but window_o.buf == NULL.");
} else
#endif
free(window_o.buf);
}
return 0;
}
/**
* @brief recvproc-function for welcome-messages
*
* @param[in] clustercmd_p Pointer to clustercmd
*
* @retval zero Successfully initialized.
* @retval non-zero Got error, while initializing.
*
*/
static int cluster_recvproc_welcome(clustercmd_t *clustercmd_p) {
debug(20, "%p", clustercmd_p);
// static time_t updatets = 0;
clustercmd_welcome_t *data_welcome_p = &clustercmd_p->data.welcome;
/*
// Is this the most recent information? Skipping if not.
if (!(data_welcome_p->updatets > updatets))
return 0;
*/
// Is the node name length in message equals to our node name length? Skipping if not.
uint32_t recv_nodename_len;
recv_nodename_len = welcome_to_node_name_len(clustercmd_p);
if (recv_nodename_len != ctx_p->cluster_nodename_len) {
debug(9, "recv_nodename_len [%i] != ctx_p->cluster_nodename_len [%i]", recv_nodename_len, ctx_p->cluster_nodename_len);
return 0;
}
// Is the node name equals to ours? Skipping if not.
if (memcmp(welcome_to_node_name(data_welcome_p), ctx_p->cluster_nodename, recv_nodename_len)) {
debug(9, "to_node_name != ctx_p->cluster_nodename");
return 0;
}
// Remembering the node that answered us
node_update_status(clustercmd_p->h.src_node_id, NODESTATUS_SEEMSONLINE);
{
char *node_name;
node_name = alloca(data_welcome_p->from_node_name_len);
memcpy(node_name, data_welcome_p->from_node_name, data_welcome_p->from_node_name_len);
node_update_name(clustercmd_p->h.src_node_id, node_name);
}
// Seems, that somebody knows our node id, remembering it.
node_id_my = clustercmd_p->h.dst_node_id;
// updatets = data_welcome_p->updatets;
return 0;
}
/**
* @brief recvproc-function for hello-messages
*
* @param[in] clustercmd_p Pointer to clustercmd
*
* @retval zero Successfully initialized.
* @retval non-zero Got error, while initializing.
*
*/
static int cluster_recvproc_hello(clustercmd_t *clustercmd_p) {
clustercmd_hello_t *data_hello_p = &clustercmd_p->data.hello;
// Sending information to the new node
{
int ret;
uint8_t node_id;
debug(15, "Preparing a welcome message for nodename \"%s\" from nodename == \"%s\"", data_hello_p->node_name, ctx_p->cluster_nodename);
size_t data_len = sizeof(clustercmd_welcome_t) + clustercmd_p->h.data_len+ctx_p->cluster_nodename_len;
clustercmd_t *answer_p = CLUSTER_ALLOCA(void, data_len);
clustercmd_welcome_t *answer_data_p = &answer_p->data.welcome;
answer_data_p->from_node_name_len = ctx_p->cluster_nodename_len;
memcpy(answer_data_p->from_node_name, ctx_p->cluster_nodename, ctx_p->cluster_nodename_len);
memcpy(welcome_to_node_name(answer_data_p), data_hello_p->node_name, clustercmd_p->h.data_len);
{
void *ptr;
char *node_name = alloca(clustercmd_p->h.data_len+1);
memcpy(node_name, data_hello_p->node_name, clustercmd_p->h.data_len+1);
ptr = g_hash_table_lookup(indexes_p->nodenames_ht, node_name);
node_id = (ptr == NULL ? NODEID_NOID : GPOINTER_TO_INT(ptr));
debug(3, "\"%s\" -> %i", node_name, node_id);
}
answer_p->h.data_len = data_len;
answer_p->h.cmd_id = CLUSTERCMDID_WELCOME;
answer_p->h.dst_node_id = node_id; // broadcast
if ((ret = cluster_send(answer_p)))
return ret;
}
return 0;
}
/**
* @brief recvproc-function for register-messages
*
* @param[in] clustercmd_p Pointer to clustercmd
*
* @retval zero Successfully initialized.
* @retval non-zero Got error, while initializing.
*
*/
static int cluster_recvproc_register(clustercmd_t *clustercmd_p) {
clustercmd_reg_t *data_register_p = &clustercmd_p->data.reg;
char *node_name = alloca(clustercmd_p->h.data_len+1);
memcpy(node_name, data_register_p->node_name, clustercmd_p->h.data_len+1);
node_update_name(clustercmd_p->h.src_node_id, node_name);
return 0;
}
extern int cluster_loop();
/**
* @brief Initializes cluster subsystem.
*
* @param[in] _ctx_p Pointer to "glob" variable, defined in main().
* @param[in] _indexes_p Pointer to "indexes" variable, defined in sync_run().
*
* @retval zero Successfully initialized.
* @retval non-zero Got error, while initializing.
*
*/
int cluster_init(ctx_t *_ctx_p, indexes_t *_indexes_p) {
int ret;
// Preventing double initializing
if (ctx_p != NULL) {
error("cluster subsystem is already initialized.");
return EALREADY;
}
// Initializing global variables, pt. 1
ctx_p = _ctx_p;
indexes_p = _indexes_p;
cluster_timeout = ctx_p->cluster_timeout;
indexes_p->nodenames_ht = g_hash_table_new_full(g_str_hash, g_str_equal, free, 0);
node_update_status(NODEID_NOID, NODESTATUS_ONLINE);
{
int i=0;
while (i < MAXNODES) {
nodeinfo[i].last_serial = -1;
i++;
}
}
// Initializing network routines
// Input socket
// Creating socket
sock_i = socket(AF_INET, SOCK_DGRAM, 0);
if(sock_i < 0) {
error("Cannot create socket for input traffic");
return errno;
}
// Enable SO_REUSEADDR to allow multiple instances of this application to receive copies
// of the multicast datagrams.
int reuse = 1;
if(setsockopt(sock_i, SOL_SOCKET, SO_REUSEADDR,(char *)&reuse, sizeof(reuse)) < 0) {
error("Got error while setsockopt()");
return errno;
}
// Binding
sa_i.sin_family = AF_INET;
sa_i.sin_port = htons(ctx_p->cluster_mcastipport);
sa_i.sin_addr.s_addr = INADDR_ANY;
if(bind(sock_i, (struct sockaddr*)&sa_i, sizeof(sa_i))) {
error("Got error while bind()");
return errno;
}
// Joining to multicast group
struct ip_mreq group;
group.imr_interface.s_addr = inet_addr(ctx_p->cluster_iface);
group.imr_multiaddr.s_addr = inet_addr(ctx_p->cluster_mcastipaddr);
if(setsockopt(sock_i, IPPROTO_IP, IP_ADD_MEMBERSHIP,
(char *)&group, sizeof(group)) < 0) {
error("Cannot setsockopt() to enter to membership %s -> %s",
ctx_p->cluster_iface, ctx_p->cluster_mcastipaddr);
return errno;
}
// Output socket
// Creating socket
sock_o = socket(AF_INET, SOCK_DGRAM, 0);
if(sock_o < 0) {
error("Cannot create socket for output traffic");
return errno;
}
// Initializing the group sockaddr structure
sa_o.sin_family = AF_INET;
sa_o.sin_port = htons(ctx_p->cluster_mcastipport);
sa_o.sin_addr.s_addr = inet_addr(ctx_p->cluster_mcastipaddr);
// Disable looping back output datagrams
{
char loopch = 0;
if(setsockopt(sock_o, IPPROTO_IP, IP_MULTICAST_LOOP, (char *)&loopch, sizeof(loopch))<0) {
error("Cannot disable loopback for output socket.");
return errno;
}
}
// Setting local interface for output traffic
{
struct in_addr addr_o;
addr_o.s_addr = inet_addr(ctx_p->cluster_iface);
if(setsockopt(sock_o, IPPROTO_IP, IP_MULTICAST_IF, &addr_o, sizeof(addr_o)) < 0) {
error("Cannot set local interface for outbound traffic");
return errno;
}
}
// Initializing another routines
cluster_io_init();
// Getting my ID in the cluster
// Trying to preserve my node_id after restart. :)
// Asking another nodes about my previous node_id
{
debug(15, "Preparing a message with my nodename == \"%s\" (%i)", ctx_p->cluster_nodename, ctx_p->cluster_nodename_len);
clustercmd_t *clustercmd_p = CLUSTER_ALLOCA(clustercmd_hello_t, ctx_p->cluster_nodename_len);
clustercmd_p->h.data_len = ctx_p->cluster_nodename_len;
memcpy(clustercmd_p->data.hello.node_name, ctx_p->cluster_nodename, clustercmd_p->h.data_len+1);
clustercmd_p->h.cmd_id = CLUSTERCMDID_HELLO;
clustercmd_p->h.dst_node_id = NODEID_NOID; // broadcast
if ((ret = cluster_send(clustercmd_p)))
return ret;
}
// Processing answers
cluster_recv_proc_set(CLUSTERCMDID_WELCOME, cluster_recvproc_welcome);
if ((ret=cluster_recv_proc(cluster_timeout)))
return ret;
// Ignore next welcome messages
cluster_recv_proc_set(CLUSTERCMDID_WELCOME, NULL);
debug(3, "After communicating with others, my node_id is %i.", node_id_my);
// Getting free node_id if nobody said us the certain value (see above).
if (node_id_my == NODEID_NOID) {
int i=0;
while (i<MAXNODES) {
if (nodeinfo[i].status == NODESTATUS_DOESNTEXIST) {
node_id_my = i;
break;
}
i++;
}
debug(3, "I was have to set my node_id to %i.", node_id_my);
}
// If there's no free id-s, then exit :(
if (node_id_my == NODEID_NOID) {
error("Cannot find free node ID. Seems, that all %i ID-s are already occupied.");
return ENOMEM;
}
// Initializing global variables, pt. 2
nodeinfo_my = &nodeinfo[node_id_my];
// Registering in the cluster
// Sending registration information
node_update_status(node_id_my, NODESTATUS_SEEMSONLINE);
{
clustercmd_t *clustercmd_p = CLUSTER_ALLOCA(clustercmd_reg_t, ctx_p->cluster_nodename_len);
clustercmd_reg_t *data_reg_p = &clustercmd_p->data.reg;
memcpy(data_reg_p->node_name, ctx_p->cluster_nodename, ctx_p->cluster_nodename_len+1);
clustercmd_p->h.data_len = ctx_p->cluster_nodename_len+1;
clustercmd_p->h.cmd_id = CLUSTERCMDID_REG;
clustercmd_p->h.dst_node_id = NODEID_NOID; // broadcast
if((ret=cluster_send(clustercmd_p)))
return ret;
}
// Setting process functions
cluster_recv_proc_set(CLUSTERCMDID_HELLO, cluster_recvproc_hello);
cluster_recv_proc_set(CLUSTERCMDID_REG, cluster_recvproc_register);
// Getting answers
if ((ret=cluster_recv_proc(cluster_timeout)))
return ret;
node_update_status(node_id_my, NODESTATUS_ONLINE);
// Running thread, that will process background communicating routines with another nodes.
// The process is based on function cluster_loop() [let's use shorthand "cluster_loop()-thread"]
ret = pthread_create(&pthread_cluster, NULL, (void *(*)(void *))cluster_loop, NULL);
return ret;
}
/**
* @brief (syncop) Sends signal to cluster_loop()-thread
*
* @param[in] signal Signal number
*
* @retval zero Successfully send the signal
* @retval non-zero Got error, while sending the signal
*
*/
static inline int cluster_signal(int signal) {
if (pthread_cluster)
return pthread_kill(pthread_cluster, signal);
return 0;
}
extern int cluster_modtime_exchange_cleanup();
/**
* @brief Antagonist of cluster_init() function. Kills cluster_loop()-thread and cleaning up
*
* @retval zero Successfully initialized
* @retval non-zero Got error, while initializing
*
*/
int cluster_deinit() {
int ret = 0;
cluster_signal(SIGTERM);
ret = pthread_join(pthread_cluster, NULL);
cluster_io_deinit();
node_update_status(NODEID_NOID, NODESTATUS_DOESNTEXIST);
#ifdef VERYPARANOID
int i=0;
#endif
while(node_count) {
#ifdef VERYPARANOID
if(i++ > MAXNODES) {
error("cluster_deinit() looped. Forcing break.");
break;
}
#endif
node_update_status(0, NODESTATUS_DOESNTEXIST);
}
close(sock_i);
close(sock_o);
#ifdef VERYPARANOID
memset(nodeinfo, 0, sizeof(nodeinfo_t) * NODES_ALLOC);
nodeinfo_my = NULL;
node_count = 0;
node_online = 0;
node_id_my = NODEID_NOID;
memset(&sa_i, 0, sizeof(sa_i));
memset(&sa_o, 0, sizeof(sa_o));
#endif
cluster_modtime_exchange_cleanup();
g_hash_table_destroy(indexes_p->nodenames_ht);
return ret;
}
/**
* @brief (syncop) Forces anothes nodes to ignore events about the file or directory
*
* @param[in] fpath Path to the file or directory
*
* @retval zero Successfully initialized
* @retval non-zero Got error, while initializing
*
*/
int cluster_lock(const char *fpath) {
return 0;
}
/**
* @brief (syncop) Forces anothes nodes to ignore events about all files and directories listed in queues of "indexes_p"
*
* @retval zero Successfully initialized
* @retval non-zero Got error, while initializing
*
*/
int cluster_lock_byindexes() {
return 0;
}
/**
* @brief (syncop) Returns events-handling on another nodes about all files and directories, locked by cluster_lock() and cluster_lock_byindexes() from this node
*
* @retval zero Successfully initialized
* @retval non-zero Got error, while initializing
*
*/
int cluster_unlock_all() {
return 0;
}
#define CLUSTER_LOOP_CHECK(a) {\
int ret = a;\
if(ret) {\
sync_term(ret);\
return ret;\
}\
}
/**
* @brief Processes background communicating routines with another nodes. cluster_init() function create a thread for this function.
*
* @retval zero Successfully initialized
* @retval non-zero Got error, while initializing
*
*/
int cluster_loop() {
int ret = 0;
sigset_t sigset_cluster;
// Ignoring SIGINT signal
sigemptyset(&sigset_cluster);
sigaddset(&sigset_cluster, SIGINT);
CLUSTER_LOOP_CHECK(pthread_sigmask(SIG_BLOCK, &sigset_cluster, NULL));
// Don't ignoring SIGTERM signal
sigemptyset(&sigset_cluster);
sigaddset(&sigset_cluster, SIGTERM);
CLUSTER_LOOP_CHECK(pthread_sigmask(SIG_UNBLOCK, &sigset_cluster, NULL));
// Starting the loop
debug(3, "cluster_loop() started.");
while (1) {
int _ret;
// Waiting for event
fd_set rfds;
FD_ZERO(&rfds);
FD_SET(sock_i, &rfds);
debug(3, "select()");
_ret = select(sock_i+1, &rfds, NULL, NULL, NULL);
// Exit if error
if ((_ret == -1) && (errno != EINTR)) {
ret = errno;
sync_term(ret);
break;
}
// Breaking the loop, if there's SIGTERM signal for this thread
debug(3, "sigpending()");
if (sigpending(&sigset_cluster))
if(sigismember(&sigset_cluster, SIGTERM))
break;
// Processing new messages
debug(3, "cluster_recv_proc()");
if ((ret=cluster_recv_proc(0))) {
sync_term(ret);
break;
}
}
debug(3, "cluster_loop() finished with exitcode %i.", ret);
return ret;
#ifdef DOXYGEN
sync_term(0);
#endif
}
/**
* @brief Updating information about modification time of a directory.
*
* @param[in] path Canonized path to updated file/dir
* @param[in] dirlevel Directory level provided by fts (man 3 fts)
* @param[in] st_mode st_mode value to detect is it directory or not (S_IFDIR or not)
*
* @retval zero Successfully initialized
* @retval non-zero Got error, while initializing
*
*/
int cluster_modtime_update(const char *path, short int dirlevel, mode_t st_mode) {
// "modtime" is incorrent name-part of function. Actually it updates "change time" (man 2 lstat64).
int ret;
// Getting relative directory level (depth)
short int dirlevel_rel = dirlevel - ctx_p->watchdir_dirlevel;
if ((st_mode & S_IFMT) == S_IFDIR)
dirlevel_rel++;
// Don't remembering information about directories with level beyond the limits
if ((dirlevel_rel > ctx_p->cluster_scan_dl_max) || (dirlevel_rel < ctx_p->cluster_hash_dl_min))
return 0;
// Getting directory/file-'s information (including "change time" aka "st_ctime")
stat64_t stat64;
ret=lstat64(path, &stat64);
if (ret) {
error("Cannot lstat64()", path);
return errno;
}
// Getting absolute directory path
const char *dirpath;
if ((st_mode & S_IFMT) == S_IFDIR) {
dirpath = path;
} else {
char *path_dup = strdup(path);
dirpath = (const char *)dirname(path_dup);
free(path_dup);
}
// Getting relative directory path
// Initializing
size_t dirpath_len = strlen(dirpath);
char *dirpath_rel_p = xmalloc(dirpath_len+1);
char *dirpath_rel = dirpath_rel_p;
const char *dirpath_rel_full = &dirpath[ctx_p->watchdirlen];
size_t dirpath_rel_full_len = dirpath_len - ctx_p->watchdirlen;
// Getting coodinate of the end (directory path is already canonized, so we can simply count number of slashes to get directory level)
int slashcount=0;
size_t dirpath_rel_end=0;
while (dirpath_rel_full[dirpath_rel_end] && (dirpath_rel_end < dirpath_rel_full_len)) {
if (dirpath_rel_full[dirpath_rel_end] == '/') {
slashcount++;
if (slashcount >= ctx_p->cluster_hash_dl_max)
break;
}
dirpath_rel_end++;
}
// Copy the required part of path to dirpath_rel
memcpy(dirpath_rel, dirpath_rel_full, dirpath_rel_end);
// Updating "st_ctime" information. We should check current value for this directory and update it only if it less or not set.
// Checking current value
char toupdate = 0;
gpointer ctime_gp = g_hash_table_lookup(nodeinfo_my->modtime_ht, dirpath_rel);
if (ctime_gp == NULL)
toupdate++;
else if (GPOINTER_TO_INT(ctime_gp) < stat64.st_ctime)
toupdate++;
// g_hash_table_replace() will replace existent information about the directory or create it if it doesn't exist.
if (toupdate)
g_hash_table_replace(nodeinfo_my->modtime_ht, strdup(dirpath_rel), GINT_TO_POINTER(stat64.st_ctime));
// Why I'm using "st_ctime" instead of "st_mtime"? Because "st_ctime" also updates on updating inode information.
return 0;
}
/**
* @brief Puts entry to list to be send to other nodes. To be called from cluster_modtime_exchange()
*
* @param[in] pushrentry_arg_p Pointer to pushentry_arg structure
*
*/
void cluster_modtime_exchange_pushentry(gpointer dir_gp, gpointer modtype_gp, void *pushentry_arg_gp) {
struct pushdoubleentry_arg *pushentry_arg_p = (struct pushdoubleentry_arg *)pushentry_arg_gp;
char *dir = (char *)dir_gp;
time_t ctime = (time_t)GPOINTER_TO_INT(modtype_gp);
size_t size = strlen(dir)+1; // TODO: strlen should be already prepared
// but not re-calculated here
if(pushentry_arg_p->allocated <= pushentry_arg_p->total) {
pushentry_arg_p->allocated += ALLOC_PORTION;
pushentry_arg_p->entry = (struct doubleentry *)
xrealloc(
(char *)pushentry_arg_p->entry,
pushentry_arg_p->allocated * sizeof(*pushentry_arg_p->entry)
);
}
pushentry_arg_p->entry[pushentry_arg_p->total].dat0 = dir;
pushentry_arg_p->entry[pushentry_arg_p->total].size0 = size;
pushentry_arg_p->entry[pushentry_arg_p->total].dat1 = (void *)ctime; // Will be problems if sizeof(time_t) > sizeof(void *)
pushentry_arg_p->entry[pushentry_arg_p->total].size1 = sizeof(ctime);
pushentry_arg_p->size += size;
pushentry_arg_p->total++;
return;
}
static struct pushdoubleentry_arg cluster_modtime_exchange_pushentry_arg = {0};
/**
* @brief Clean up after the last run of cluster_modtime_exchange.
*
* @retval zero Successfully initialized
* @retval non-zero Got error, while initializing
*
*/
int cluster_modtime_exchange_cleanup() {
struct pushdoubleentry_arg *pushentry_arg_p = &cluster_modtime_exchange_pushentry_arg;
int i=0;
while(i < pushentry_arg_p->allocated) {
if(pushentry_arg_p->entry[i].alloc0)
free(pushentry_arg_p->entry[i].dat0);
if(pushentry_arg_p->entry[i].alloc1)
free(pushentry_arg_p->entry[i].dat1);
i++;
}
free(pushentry_arg_p->entry);
#ifdef VERYPARANOID
memset(pushentry_arg_p, 0, sizeof(*pushentry_arg_p));
#endif
return 0;
}
/**
* @brief Exchanging with "modtime_ht"-s to be able to compare them.
*
* @retval zero Successfully initialized
* @retval non-zero Got error, while initializing
*
*/
int cluster_modtime_exchange() {
struct pushdoubleentry_arg *pushentry_arg_p = &cluster_modtime_exchange_pushentry_arg;
// Getting hash table entries
pushentry_arg_p->size=0;
pushentry_arg_p->total=0;
g_hash_table_foreach(nodeinfo_my->modtime_ht, cluster_modtime_exchange_pushentry, (void *)pushentry_arg_p);
if(!pushentry_arg_p->total) {
// !!!
}
// Calculating required RAM to compile clustercmd
size_t toalloc = 0;
int i = 0;
while(i < pushentry_arg_p->total) {
toalloc += 4; // for size header
toalloc += pushentry_arg_p->entry[i].size0; // for path
toalloc += pushentry_arg_p->entry[i].size1; // for ctime
}
// Allocating space for the clustercmd
clustercmd_t *clustercmd_p = (clustercmd_t *)xmalloc(sizeof(clustercmdhdr_t) + toalloc);
memset(clustercmd_p, 0, sizeof(clustercmdhdr_t));
// Setting up clustercmd
clustercmd_p->h.dst_node_id = NODEID_NOID;
clustercmd_p->h.cmd_id = CLUSTERCMDID_HT_EXCH;
clustercmd_p->h.data_len = toalloc;
// Filing clustercmd with hash-table entriyes
i = 0;
clustercmd_ht_exch_t *clustercmd_ht_exch_p = &clustercmd_p->data.ht_exch;
while(i < pushentry_arg_p->total) {
// Setting the data
clustercmd_ht_exch_p->ctime = (time_t)pushentry_arg_p->entry[i].dat1;
clustercmd_ht_exch_p->path_length = (time_t)pushentry_arg_p->entry[i].size0;
memcpy(
clustercmd_ht_exch_p->path,
pushentry_arg_p->entry[i].dat0,
clustercmd_ht_exch_p->path_length
);
// Pointing to space for next entry:
size_t offset = sizeof(clustercmd_ht_exch_t)-1+pushentry_arg_p->entry[i].size0;
clustercmd_ht_exch_p = (clustercmd_ht_exch_t *)
(&((char *) clustercmd_ht_exch_p)[offset] );
}
// Sending
cluster_send(clustercmd_p);
// Cleanup
free(clustercmd_p);
return 0;
}
/**
* @brief (syncop) Syncing file tree with another nodes with using of directories' modification time as a recent-detector.
*
* @param[in] dirpath Path to the directory
*
* @retval zero Successfully initialized
* @retval non-zero Got error, while initializing
*
*/
int cluster_initialsync() {
cluster_modtime_exchange();
return 0;
}
/**
* @brief (syncop) "Captures" right to update the file or directory to another nodes. It just removes events about the file of directory from another nodes
*
* @param[in] dirpath Path to the directory
*
* @retval zero Successfully initialized
* @retval non-zero Got error, while initializing
*
*/
int cluster_capture(const char *path) {
return 0;
}
#endif