forked from FRRouting/frr
-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathpim_mlag.c
1101 lines (935 loc) · 29.9 KB
/
pim_mlag.c
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
/*
* This is an implementation of PIM MLAG Functionality
*
* Module name: PIM MLAG
*
* Author: sathesh Kumar karra <sathk@cumulusnetworks.com>
*
* Copyright (C) 2019 Cumulus Networks http://www.cumulusnetworks.com
*
* 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 2 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; see the file COPYING; if not, write to the Free Software
* Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include <zebra.h>
#include "pimd.h"
#include "pim_mlag.h"
#include "pim_upstream.h"
#include "pim_vxlan.h"
extern struct zclient *zclient;
#define PIM_MLAG_METADATA_LEN 4
/*********************ACtual Data processing *****************************/
/* TBD: There can be duplicate updates to FIB***/
#define PIM_MLAG_ADD_OIF_TO_OIL(ch, ch_oil) \
do { \
if (PIM_DEBUG_MLAG) \
zlog_debug( \
"%s: add Dual-active Interface to %s " \
"to oil:%s", \
__func__, ch->interface->name, ch->sg_str); \
pim_channel_update_oif_mute(ch_oil, ch->interface->info); \
} while (0)
#define PIM_MLAG_DEL_OIF_TO_OIL(ch, ch_oil) \
do { \
if (PIM_DEBUG_MLAG) \
zlog_debug( \
"%s: del Dual-active Interface to %s " \
"to oil:%s", \
__func__, ch->interface->name, ch->sg_str); \
pim_channel_update_oif_mute(ch_oil, ch->interface->info); \
} while (0)
static void pim_mlag_calculate_df_for_ifchannels(struct pim_upstream *up,
bool is_df)
{
struct listnode *chnode;
struct listnode *chnextnode;
struct pim_ifchannel *ch;
struct pim_interface *pim_ifp = NULL;
struct channel_oil *ch_oil = NULL;
ch_oil = (up) ? up->channel_oil : NULL;
if (!ch_oil)
return;
if (PIM_DEBUG_MLAG)
zlog_debug("%s: Calculating DF for Dual active if-channel%s",
__func__, up->sg_str);
for (ALL_LIST_ELEMENTS(up->ifchannels, chnode, chnextnode, ch)) {
pim_ifp = (ch->interface) ? ch->interface->info : NULL;
if (!pim_ifp || !PIM_I_am_DualActive(pim_ifp))
continue;
if (is_df)
PIM_MLAG_ADD_OIF_TO_OIL(ch, ch_oil);
else
PIM_MLAG_DEL_OIF_TO_OIL(ch, ch_oil);
}
}
static void pim_mlag_inherit_mlag_flags(struct pim_upstream *up, bool is_df)
{
struct listnode *listnode;
struct pim_upstream *child;
struct listnode *chnode;
struct listnode *chnextnode;
struct pim_ifchannel *ch;
struct pim_interface *pim_ifp = NULL;
struct channel_oil *ch_oil = NULL;
if (PIM_DEBUG_MLAG)
zlog_debug("%s: Updating DF for uptream:%s children", __func__,
up->sg_str);
for (ALL_LIST_ELEMENTS(up->ifchannels, chnode, chnextnode, ch)) {
pim_ifp = (ch->interface) ? ch->interface->info : NULL;
if (!pim_ifp || !PIM_I_am_DualActive(pim_ifp))
continue;
for (ALL_LIST_ELEMENTS_RO(up->sources, listnode, child)) {
if (PIM_DEBUG_MLAG)
zlog_debug("%s: Updating DF for child:%s",
__func__, child->sg_str);
ch_oil = (child) ? child->channel_oil : NULL;
if (!ch_oil)
continue;
if (is_df)
PIM_MLAG_ADD_OIF_TO_OIL(ch, ch_oil);
else
PIM_MLAG_DEL_OIF_TO_OIL(ch, ch_oil);
}
}
}
/******************************* pim upstream sync **************************/
/* Update DF role for the upstream entry and return true on role change */
bool pim_mlag_up_df_role_update(struct pim_instance *pim,
struct pim_upstream *up, bool is_df, const char *reason)
{
struct channel_oil *c_oil = up->channel_oil;
bool old_is_df = !PIM_UPSTREAM_FLAG_TEST_MLAG_NON_DF(up->flags);
struct pim_interface *vxlan_ifp;
if (is_df == old_is_df) {
if (PIM_DEBUG_MLAG)
zlog_debug(
"%s: Ignoring Role update for %s, since no change",
__func__, up->sg_str);
return false;
}
if (PIM_DEBUG_MLAG)
zlog_debug("local MLAG mroute %s role changed to %s based on %s",
up->sg_str, is_df ? "df" : "non-df", reason);
if (is_df)
PIM_UPSTREAM_FLAG_UNSET_MLAG_NON_DF(up->flags);
else
PIM_UPSTREAM_FLAG_SET_MLAG_NON_DF(up->flags);
/*
* This Upstream entry synced to peer Because of Dual-active
* Interface configuration
*/
if (PIM_UPSTREAM_FLAG_TEST_MLAG_INTERFACE(up->flags)) {
pim_mlag_inherit_mlag_flags(up, is_df);
pim_mlag_calculate_df_for_ifchannels(up, is_df);
}
/* If the DF role has changed check if ipmr-lo needs to be
* muted/un-muted. Active-Active devices and vxlan termination
* devices (ipmr-lo) are suppressed on the non-DF.
* This may leave the mroute with the empty OIL in which case the
* the forwarding entry's sole purpose is to just blackhole the flow
* headed to the switch.
*/
if (c_oil) {
vxlan_ifp = pim_vxlan_get_term_ifp(pim);
if (vxlan_ifp)
pim_channel_update_oif_mute(c_oil, vxlan_ifp);
}
/* If DF role changed on a (*,G) termination mroute update the
* associated DF role on the inherited (S,G) entries
*/
if ((up->sg.src.s_addr == INADDR_ANY) &&
PIM_UPSTREAM_FLAG_TEST_MLAG_VXLAN(up->flags))
pim_vxlan_inherit_mlag_flags(pim, up, true /* inherit */);
return true;
}
/* Run per-upstream entry DF election and return true on role change */
static bool pim_mlag_up_df_role_elect(struct pim_instance *pim,
struct pim_upstream *up)
{
bool is_df;
uint32_t peer_cost;
uint32_t local_cost;
bool rv;
if (!pim_up_mlag_is_local(up))
return false;
/* We are yet to rx a status update from the local MLAG daemon so
* we will assume DF status.
*/
if (!(router->mlag_flags & PIM_MLAGF_STATUS_RXED))
return pim_mlag_up_df_role_update(pim, up,
true /*is_df*/, "mlagd-down");
/* If not connected to peer assume DF role on the MLAG primary
* switch (and non-DF on the secondary switch.
*/
if (!(router->mlag_flags & PIM_MLAGF_PEER_CONN_UP)) {
is_df = (router->mlag_role == MLAG_ROLE_PRIMARY) ? true : false;
return pim_mlag_up_df_role_update(pim, up,
is_df, "peer-down");
}
/* If MLAG peer session is up but zebra is down on the peer
* assume DF role.
*/
if (!(router->mlag_flags & PIM_MLAGF_PEER_ZEBRA_UP))
return pim_mlag_up_df_role_update(pim, up,
true /*is_df*/, "zebra-down");
/* If we are connected to peer switch but don't have a mroute
* from it we have to assume non-DF role to avoid duplicates.
* Note: When the peer connection comes up we wait for initial
* replay to complete before moving "strays" i.e. local-mlag-mroutes
* without a peer reference to non-df role.
*/
if (!PIM_UPSTREAM_FLAG_TEST_MLAG_PEER(up->flags))
return pim_mlag_up_df_role_update(pim, up,
false /*is_df*/, "no-peer-mroute");
/* switch with the lowest RPF cost wins. if both switches have the same
* cost MLAG role is used as a tie breaker (MLAG primary wins).
*/
peer_cost = up->mlag.peer_mrib_metric;
local_cost = pim_up_mlag_local_cost(up);
if (local_cost == peer_cost) {
is_df = (router->mlag_role == MLAG_ROLE_PRIMARY) ? true : false;
rv = pim_mlag_up_df_role_update(pim, up, is_df, "equal-cost");
} else {
is_df = (local_cost < peer_cost) ? true : false;
rv = pim_mlag_up_df_role_update(pim, up, is_df, "cost");
}
return rv;
}
/* Handle upstream entry add from the peer MLAG switch -
* - if a local entry doesn't exist one is created with reference
* _MLAG_PEER
* - if a local entry exists and has a MLAG OIF DF election is run.
* the non-DF switch stop forwarding traffic to MLAG devices.
*/
static void pim_mlag_up_peer_add(struct mlag_mroute_add *msg)
{
struct pim_upstream *up;
struct pim_instance *pim;
int flags = 0;
struct prefix_sg sg;
struct vrf *vrf;
char sg_str[PIM_SG_LEN];
memset(&sg, 0, sizeof(struct prefix_sg));
sg.src.s_addr = htonl(msg->source_ip);
sg.grp.s_addr = htonl(msg->group_ip);
if (PIM_DEBUG_MLAG)
pim_str_sg_set(&sg, sg_str);
if (PIM_DEBUG_MLAG)
zlog_debug("peer MLAG mroute add %s:%s cost %d",
msg->vrf_name, sg_str, msg->cost_to_rp);
/* XXX - this is not correct. we MUST cache updates to avoid losing
* an entry because of race conditions with the peer switch.
*/
vrf = vrf_lookup_by_name(msg->vrf_name);
if (!vrf) {
if (PIM_DEBUG_MLAG)
zlog_debug("peer MLAG mroute add failed %s:%s; no vrf",
msg->vrf_name, sg_str);
return;
}
pim = vrf->info;
up = pim_upstream_find(pim, &sg);
if (up) {
/* upstream already exists; create peer reference if it
* doesn't already exist.
*/
if (!PIM_UPSTREAM_FLAG_TEST_MLAG_PEER(up->flags))
pim_upstream_ref(up, PIM_UPSTREAM_FLAG_MASK_MLAG_PEER,
__func__);
} else {
PIM_UPSTREAM_FLAG_SET_MLAG_PEER(flags);
up = pim_upstream_add(pim, &sg, NULL /*iif*/, flags, __func__,
NULL /*if_ch*/);
if (!up) {
if (PIM_DEBUG_MLAG)
zlog_debug("peer MLAG mroute add failed %s:%s",
vrf->name, sg_str);
return;
}
}
up->mlag.peer_mrib_metric = msg->cost_to_rp;
pim_mlag_up_df_role_elect(pim, up);
}
/* Handle upstream entry del from the peer MLAG switch -
* - peer reference is removed. this can result in the upstream
* being deleted altogether.
* - if a local entry continues to exisy and has a MLAG OIF DF election
* is re-run (at the end of which the local entry will be the DF).
*/
static struct pim_upstream *pim_mlag_up_peer_deref(struct pim_instance *pim,
struct pim_upstream *up)
{
if (!PIM_UPSTREAM_FLAG_TEST_MLAG_PEER(up->flags))
return up;
PIM_UPSTREAM_FLAG_UNSET_MLAG_PEER(up->flags);
up = pim_upstream_del(pim, up, __func__);
if (up)
pim_mlag_up_df_role_elect(pim, up);
return up;
}
static void pim_mlag_up_peer_del(struct mlag_mroute_del *msg)
{
struct pim_upstream *up;
struct pim_instance *pim;
struct prefix_sg sg;
struct vrf *vrf;
char sg_str[PIM_SG_LEN];
memset(&sg, 0, sizeof(struct prefix_sg));
sg.src.s_addr = htonl(msg->source_ip);
sg.grp.s_addr = htonl(msg->group_ip);
if (PIM_DEBUG_MLAG)
pim_str_sg_set(&sg, sg_str);
if (PIM_DEBUG_MLAG)
zlog_debug("peer MLAG mroute del %s:%s", msg->vrf_name,
sg_str);
vrf = vrf_lookup_by_name(msg->vrf_name);
if (!vrf) {
if (PIM_DEBUG_MLAG)
zlog_debug("peer MLAG mroute del skipped %s:%s; no vrf",
msg->vrf_name, sg_str);
return;
}
pim = vrf->info;
up = pim_upstream_find(pim, &sg);
if (!up) {
if (PIM_DEBUG_MLAG)
zlog_debug("peer MLAG mroute del skipped %s:%s; no up",
vrf->name, sg_str);
return;
}
(void)pim_mlag_up_peer_deref(pim, up);
}
/* When we lose connection to the local MLAG daemon we can drop all peer
* references.
*/
static void pim_mlag_up_peer_del_all(void)
{
struct list *temp = list_new();
struct pim_upstream *up;
struct vrf *vrf;
struct pim_instance *pim;
/*
* So why these gyrations?
* pim->upstream_head has the list of *,G and S,G
* that are in the system. The problem of course
* is that it is an ordered list:
* (*,G1) -> (S1,G1) -> (S2,G2) -> (S3, G2) -> (*,G2) -> (S1,G2)
* And the *,G1 has pointers to S1,G1 and S2,G1
* if we delete *,G1 then we have a situation where
* S1,G1 and S2,G2 can be deleted as well. Then a
* simple ALL_LIST_ELEMENTS will have the next listnode
* pointer become invalid and we crash.
* So let's grab the list of MLAG_PEER upstreams
* add a refcount put on another list and delete safely
*/
RB_FOREACH(vrf, vrf_name_head, &vrfs_by_name) {
pim = vrf->info;
frr_each (rb_pim_upstream, &pim->upstream_head, up) {
if (!PIM_UPSTREAM_FLAG_TEST_MLAG_PEER(up->flags))
continue;
listnode_add(temp, up);
/*
* Add a reference since we are adding to this
* list for deletion
*/
up->ref_count++;
}
while (temp->count) {
up = listnode_head(temp);
listnode_delete(temp, up);
up = pim_mlag_up_peer_deref(pim, up);
/*
* This is the deletion of the reference added
* above
*/
if (up)
pim_upstream_del(pim, up, __func__);
}
}
list_delete(&temp);
}
/* Send upstream entry to the local MLAG daemon (which will subsequently
* send it to the peer MLAG switch).
*/
static void pim_mlag_up_local_add_send(struct pim_instance *pim,
struct pim_upstream *up)
{
struct stream *s = NULL;
struct vrf *vrf = pim->vrf;
if (!(router->mlag_flags & PIM_MLAGF_LOCAL_CONN_UP))
return;
s = stream_new(sizeof(struct mlag_mroute_add) + PIM_MLAG_METADATA_LEN);
if (!s)
return;
if (PIM_DEBUG_MLAG)
zlog_debug("local MLAG mroute add %s:%s",
vrf->name, up->sg_str);
++router->mlag_stats.msg.mroute_add_tx;
stream_putl(s, MLAG_MROUTE_ADD);
stream_put(s, vrf->name, VRF_NAMSIZ);
stream_putl(s, ntohl(up->sg.src.s_addr));
stream_putl(s, ntohl(up->sg.grp.s_addr));
stream_putl(s, pim_up_mlag_local_cost(up));
/* XXX - who is addding*/
stream_putl(s, MLAG_OWNER_VXLAN);
/* XXX - am_i_DR field should be removed */
stream_putc(s, false);
stream_putc(s, !(PIM_UPSTREAM_FLAG_TEST_MLAG_NON_DF(up->flags)));
stream_putl(s, vrf->vrf_id);
/* XXX - this field is a No-op for VXLAN*/
stream_put(s, NULL, INTERFACE_NAMSIZ);
stream_fifo_push_safe(router->mlag_fifo, s);
pim_mlag_signal_zpthread();
}
static void pim_mlag_up_local_del_send(struct pim_instance *pim,
struct pim_upstream *up)
{
struct stream *s = NULL;
struct vrf *vrf = pim->vrf;
if (!(router->mlag_flags & PIM_MLAGF_LOCAL_CONN_UP))
return;
s = stream_new(sizeof(struct mlag_mroute_del) + PIM_MLAG_METADATA_LEN);
if (!s)
return;
if (PIM_DEBUG_MLAG)
zlog_debug("local MLAG mroute del %s:%s",
vrf->name, up->sg_str);
++router->mlag_stats.msg.mroute_del_tx;
stream_putl(s, MLAG_MROUTE_DEL);
stream_put(s, vrf->name, VRF_NAMSIZ);
stream_putl(s, ntohl(up->sg.src.s_addr));
stream_putl(s, ntohl(up->sg.grp.s_addr));
/* XXX - who is adding */
stream_putl(s, MLAG_OWNER_VXLAN);
stream_putl(s, vrf->vrf_id);
/* XXX - this field is a No-op for VXLAN */
stream_put(s, NULL, INTERFACE_NAMSIZ);
/* XXX - is this the the most optimal way to do things */
stream_fifo_push_safe(router->mlag_fifo, s);
pim_mlag_signal_zpthread();
}
/* Called when a local upstream entry is created or if it's cost changes */
void pim_mlag_up_local_add(struct pim_instance *pim,
struct pim_upstream *up)
{
pim_mlag_up_df_role_elect(pim, up);
/* XXX - need to add some dup checks here */
pim_mlag_up_local_add_send(pim, up);
}
/* Called when local MLAG reference is removed from an upstream entry */
void pim_mlag_up_local_del(struct pim_instance *pim,
struct pim_upstream *up)
{
pim_mlag_up_df_role_elect(pim, up);
pim_mlag_up_local_del_send(pim, up);
}
/* When connection to local MLAG daemon is established all the local
* MLAG upstream entries are replayed to it.
*/
static void pim_mlag_up_local_replay(void)
{
struct pim_upstream *up;
struct vrf *vrf;
struct pim_instance *pim;
RB_FOREACH(vrf, vrf_name_head, &vrfs_by_name) {
pim = vrf->info;
frr_each (rb_pim_upstream, &pim->upstream_head, up) {
if (pim_up_mlag_is_local(up))
pim_mlag_up_local_add_send(pim, up);
}
}
}
/* on local/peer mlag connection and role changes the DF status needs
* to be re-evaluated
*/
static void pim_mlag_up_local_reeval(bool mlagd_send, const char *reason_code)
{
struct pim_upstream *up;
struct vrf *vrf;
struct pim_instance *pim;
if (PIM_DEBUG_MLAG)
zlog_debug("%s re-run DF election because of %s",
__func__, reason_code);
RB_FOREACH(vrf, vrf_name_head, &vrfs_by_name) {
pim = vrf->info;
frr_each (rb_pim_upstream, &pim->upstream_head, up) {
if (!pim_up_mlag_is_local(up))
continue;
/* if role changes re-send to peer */
if (pim_mlag_up_df_role_elect(pim, up) &&
mlagd_send)
pim_mlag_up_local_add_send(pim, up);
}
}
}
/*****************PIM Actions for MLAG state changes**********************/
/* notify the anycast VTEP component about state changes */
static inline void pim_mlag_vxlan_state_update(void)
{
bool enable = !!(router->mlag_flags & PIM_MLAGF_STATUS_RXED);
bool peer_state = !!(router->mlag_flags & PIM_MLAGF_PEER_CONN_UP);
pim_vxlan_mlag_update(enable, peer_state, router->mlag_role,
router->peerlink_rif_p, &router->local_vtep_ip);
}
/**************End of PIM Actions for MLAG State changes******************/
/********************API to process PIM MLAG Data ************************/
static void pim_mlag_process_mlagd_state_change(struct mlag_status msg)
{
bool role_chg = false;
bool state_chg = false;
bool notify_vxlan = false;
struct interface *peerlink_rif_p;
char buf[MLAG_ROLE_STRSIZE];
if (PIM_DEBUG_MLAG)
zlog_debug("%s: msg dump: my_role: %s, peer_state: %s",
__func__,
mlag_role2str(msg.my_role, buf, sizeof(buf)),
(msg.peer_state == MLAG_STATE_RUNNING ? "RUNNING"
: "DOWN"));
if (!(router->mlag_flags & PIM_MLAGF_LOCAL_CONN_UP)) {
if (PIM_DEBUG_MLAG)
zlog_debug("%s: msg ignored mlagd process state down",
__func__);
return;
}
++router->mlag_stats.msg.mlag_status_updates;
/* evaluate the changes first */
if (router->mlag_role != msg.my_role) {
role_chg = true;
notify_vxlan = true;
router->mlag_role = msg.my_role;
}
strlcpy(router->peerlink_rif, msg.peerlink_rif,
sizeof(router->peerlink_rif));
/* XXX - handle the case where we may rx the interface name from the
* MLAG daemon before we get the interface from zebra.
*/
peerlink_rif_p = if_lookup_by_name(router->peerlink_rif, VRF_DEFAULT);
if (router->peerlink_rif_p != peerlink_rif_p) {
router->peerlink_rif_p = peerlink_rif_p;
notify_vxlan = true;
}
if (msg.peer_state == MLAG_STATE_RUNNING) {
if (!(router->mlag_flags & PIM_MLAGF_PEER_CONN_UP)) {
state_chg = true;
notify_vxlan = true;
router->mlag_flags |= PIM_MLAGF_PEER_CONN_UP;
}
router->connected_to_mlag = true;
} else {
if (router->mlag_flags & PIM_MLAGF_PEER_CONN_UP) {
++router->mlag_stats.peer_session_downs;
state_chg = true;
notify_vxlan = true;
router->mlag_flags &= ~PIM_MLAGF_PEER_CONN_UP;
}
router->connected_to_mlag = false;
}
/* apply the changes */
/* when connection to mlagd comes up we hold send mroutes till we have
* rxed the status and had a chance to re-valuate DF state
*/
if (!(router->mlag_flags & PIM_MLAGF_STATUS_RXED)) {
router->mlag_flags |= PIM_MLAGF_STATUS_RXED;
pim_mlag_vxlan_state_update();
/* on session up re-eval DF status */
pim_mlag_up_local_reeval(false /*mlagd_send*/, "mlagd_up");
/* replay all the upstream entries to the local MLAG daemon */
pim_mlag_up_local_replay();
return;
}
if (notify_vxlan)
pim_mlag_vxlan_state_update();
if (state_chg) {
if (!(router->mlag_flags & PIM_MLAGF_PEER_CONN_UP))
/* when a connection goes down the primary takes over
* DF role for all entries
*/
pim_mlag_up_local_reeval(true /*mlagd_send*/,
"peer_down");
else
/* XXX - when session comes up we need to wait for
* PEER_REPLAY_DONE before running re-election on
* local-mlag entries that are missing peer reference
*/
pim_mlag_up_local_reeval(true /*mlagd_send*/,
"peer_up");
} else if (role_chg) {
/* MLAG role changed without a state change */
pim_mlag_up_local_reeval(true /*mlagd_send*/, "role_chg");
}
}
static void pim_mlag_process_peer_frr_state_change(struct mlag_frr_status msg)
{
if (PIM_DEBUG_MLAG)
zlog_debug(
"%s: msg dump: peer_frr_state: %s", __func__,
(msg.frr_state == MLAG_FRR_STATE_UP ? "UP" : "DOWN"));
if (!(router->mlag_flags & PIM_MLAGF_LOCAL_CONN_UP)) {
if (PIM_DEBUG_MLAG)
zlog_debug("%s: msg ignored mlagd process state down",
__func__);
return;
}
++router->mlag_stats.msg.peer_zebra_status_updates;
/* evaluate the changes first */
if (msg.frr_state == MLAG_FRR_STATE_UP) {
if (!(router->mlag_flags & PIM_MLAGF_PEER_ZEBRA_UP)) {
router->mlag_flags |= PIM_MLAGF_PEER_ZEBRA_UP;
/* XXX - when peer zebra comes up we need to wait for
* for some time to let the peer setup MDTs before
* before relinquishing DF status
*/
pim_mlag_up_local_reeval(true /*mlagd_send*/,
"zebra_up");
}
} else {
if (router->mlag_flags & PIM_MLAGF_PEER_ZEBRA_UP) {
++router->mlag_stats.peer_zebra_downs;
router->mlag_flags &= ~PIM_MLAGF_PEER_ZEBRA_UP;
/* when a peer zebra goes down we assume DF role */
pim_mlag_up_local_reeval(true /*mlagd_send*/,
"zebra_down");
}
}
}
static void pim_mlag_process_vxlan_update(struct mlag_vxlan *msg)
{
char addr_buf1[INET_ADDRSTRLEN];
char addr_buf2[INET_ADDRSTRLEN];
uint32_t local_ip;
if (!(router->mlag_flags & PIM_MLAGF_LOCAL_CONN_UP)) {
if (PIM_DEBUG_MLAG)
zlog_debug("%s: msg ignored mlagd process state down",
__func__);
return;
}
++router->mlag_stats.msg.vxlan_updates;
router->anycast_vtep_ip.s_addr = htonl(msg->anycast_ip);
local_ip = htonl(msg->local_ip);
if (router->local_vtep_ip.s_addr != local_ip) {
router->local_vtep_ip.s_addr = local_ip;
pim_mlag_vxlan_state_update();
}
if (PIM_DEBUG_MLAG) {
inet_ntop(AF_INET, &router->local_vtep_ip,
addr_buf1, INET_ADDRSTRLEN);
inet_ntop(AF_INET, &router->anycast_vtep_ip,
addr_buf2, INET_ADDRSTRLEN);
zlog_debug("%s: msg dump: local-ip:%s, anycast-ip:%s",
__func__, addr_buf1, addr_buf2);
}
}
static void pim_mlag_process_mroute_add(struct mlag_mroute_add msg)
{
if (PIM_DEBUG_MLAG) {
struct prefix_sg sg;
sg.grp.s_addr = ntohl(msg.group_ip);
sg.src.s_addr = ntohl(msg.source_ip);
zlog_debug(
"%s: msg dump: vrf_name: %s, s.ip: 0x%x, g.ip: 0x%x (%pSG4) cost: %u",
__func__, msg.vrf_name, msg.source_ip, msg.group_ip,
&sg, msg.cost_to_rp);
zlog_debug(
"(%pSG4)owner_id: %d, DR: %d, Dual active: %d, vrf_id: 0x%x intf_name: %s",
&sg, msg.owner_id, msg.am_i_dr, msg.am_i_dual_active,
msg.vrf_id, msg.intf_name);
}
if (!(router->mlag_flags & PIM_MLAGF_LOCAL_CONN_UP)) {
if (PIM_DEBUG_MLAG)
zlog_debug("%s: msg ignored mlagd process state down",
__func__);
return;
}
++router->mlag_stats.msg.mroute_add_rx;
pim_mlag_up_peer_add(&msg);
}
static void pim_mlag_process_mroute_del(struct mlag_mroute_del msg)
{
if (PIM_DEBUG_MLAG) {
struct prefix_sg sg;
sg.grp.s_addr = ntohl(msg.group_ip);
sg.src.s_addr = ntohl(msg.source_ip);
zlog_debug(
"%s: msg dump: vrf_name: %s, s.ip: 0x%x, g.ip: 0x%x(%pSG4)",
__func__, msg.vrf_name, msg.source_ip, msg.group_ip,
&sg);
zlog_debug("(%pSG4)owner_id: %d, vrf_id: 0x%x intf_name: %s",
&sg, msg.owner_id, msg.vrf_id, msg.intf_name);
}
if (!(router->mlag_flags & PIM_MLAGF_LOCAL_CONN_UP)) {
if (PIM_DEBUG_MLAG)
zlog_debug("%s: msg ignored mlagd process state down",
__func__);
return;
}
++router->mlag_stats.msg.mroute_del_rx;
pim_mlag_up_peer_del(&msg);
}
int pim_zebra_mlag_handle_msg(struct stream *s, int len)
{
struct mlag_msg mlag_msg;
char buf[80];
int rc = 0;
size_t length;
rc = mlag_lib_decode_mlag_hdr(s, &mlag_msg, &length);
if (rc)
return (rc);
if (PIM_DEBUG_MLAG)
zlog_debug("%s: Received msg type: %s length: %d, bulk_cnt: %d",
__func__,
mlag_lib_msgid_to_str(mlag_msg.msg_type, buf,
sizeof(buf)),
mlag_msg.data_len, mlag_msg.msg_cnt);
switch (mlag_msg.msg_type) {
case MLAG_STATUS_UPDATE: {
struct mlag_status msg;
rc = mlag_lib_decode_mlag_status(s, &msg);
if (rc)
return (rc);
pim_mlag_process_mlagd_state_change(msg);
} break;
case MLAG_PEER_FRR_STATUS: {
struct mlag_frr_status msg;
rc = mlag_lib_decode_frr_status(s, &msg);
if (rc)
return (rc);
pim_mlag_process_peer_frr_state_change(msg);
} break;
case MLAG_VXLAN_UPDATE: {
struct mlag_vxlan msg;
rc = mlag_lib_decode_vxlan_update(s, &msg);
if (rc)
return rc;
pim_mlag_process_vxlan_update(&msg);
} break;
case MLAG_MROUTE_ADD: {
struct mlag_mroute_add msg;
rc = mlag_lib_decode_mroute_add(s, &msg, &length);
if (rc)
return (rc);
pim_mlag_process_mroute_add(msg);
} break;
case MLAG_MROUTE_DEL: {
struct mlag_mroute_del msg;
rc = mlag_lib_decode_mroute_del(s, &msg, &length);
if (rc)
return (rc);
pim_mlag_process_mroute_del(msg);
} break;
case MLAG_MROUTE_ADD_BULK: {
struct mlag_mroute_add msg;
int i;
for (i = 0; i < mlag_msg.msg_cnt; i++) {
rc = mlag_lib_decode_mroute_add(s, &msg, &length);
if (rc)
return (rc);
pim_mlag_process_mroute_add(msg);
}
} break;
case MLAG_MROUTE_DEL_BULK: {
struct mlag_mroute_del msg;
int i;
for (i = 0; i < mlag_msg.msg_cnt; i++) {
rc = mlag_lib_decode_mroute_del(s, &msg, &length);
if (rc)
return (rc);
pim_mlag_process_mroute_del(msg);
}
} break;
default:
break;
}
return 0;
}
/****************End of PIM Mesasge processing handler********************/
int pim_zebra_mlag_process_up(void)
{
if (PIM_DEBUG_MLAG)
zlog_debug("%s: Received Process-Up from Mlag", __func__);
/*
* Incase of local MLAG restart, PIM needs to replay all the data
* since MLAG is empty.
*/
router->connected_to_mlag = true;
router->mlag_flags |= PIM_MLAGF_LOCAL_CONN_UP;
return 0;
}
static void pim_mlag_param_reset(void)
{
/* reset the cached params and stats */
router->mlag_flags &= ~(PIM_MLAGF_STATUS_RXED |
PIM_MLAGF_LOCAL_CONN_UP |
PIM_MLAGF_PEER_CONN_UP |
PIM_MLAGF_PEER_ZEBRA_UP);
router->local_vtep_ip.s_addr = INADDR_ANY;
router->anycast_vtep_ip.s_addr = INADDR_ANY;
router->mlag_role = MLAG_ROLE_NONE;
memset(&router->mlag_stats.msg, 0, sizeof(router->mlag_stats.msg));
router->peerlink_rif[0] = '\0';
}
int pim_zebra_mlag_process_down(void)
{
if (PIM_DEBUG_MLAG)
zlog_debug("%s: Received Process-Down from Mlag", __func__);
/* Local CLAG is down, reset peer data and forward the traffic if
* we are DR
*/
if (router->mlag_flags & PIM_MLAGF_PEER_CONN_UP)
++router->mlag_stats.peer_session_downs;
if (router->mlag_flags & PIM_MLAGF_PEER_ZEBRA_UP)
++router->mlag_stats.peer_zebra_downs;
router->connected_to_mlag = false;
pim_mlag_param_reset();
/* on mlagd session down re-eval DF status */
pim_mlag_up_local_reeval(false /*mlagd_send*/, "mlagd_down");
/* flush all peer references */
pim_mlag_up_peer_del_all();
/* notify the vxlan component */
pim_mlag_vxlan_state_update();
return 0;
}
static int pim_mlag_register_handler(struct thread *thread)
{
uint32_t bit_mask = 0;
if (!zclient)
return -1;
SET_FLAG(bit_mask, (1 << MLAG_STATUS_UPDATE));
SET_FLAG(bit_mask, (1 << MLAG_MROUTE_ADD));
SET_FLAG(bit_mask, (1 << MLAG_MROUTE_DEL));
SET_FLAG(bit_mask, (1 << MLAG_DUMP));
SET_FLAG(bit_mask, (1 << MLAG_MROUTE_ADD_BULK));
SET_FLAG(bit_mask, (1 << MLAG_MROUTE_DEL_BULK));
SET_FLAG(bit_mask, (1 << MLAG_PIM_CFG_DUMP));
SET_FLAG(bit_mask, (1 << MLAG_VXLAN_UPDATE));
SET_FLAG(bit_mask, (1 << MLAG_PEER_FRR_STATUS));
if (PIM_DEBUG_MLAG)
zlog_debug("%s: Posting Client Register to MLAG mask: 0x%x",
__func__, bit_mask);
zclient_send_mlag_register(zclient, bit_mask);
return 0;
}
void pim_mlag_register(void)
{
if (router->mlag_process_register)
return;
router->mlag_process_register = true;
thread_add_event(router->master, pim_mlag_register_handler, NULL, 0,
NULL);
}
static int pim_mlag_deregister_handler(struct thread *thread)
{
if (!zclient)
return -1;
if (PIM_DEBUG_MLAG)
zlog_debug("%s: Posting Client De-Register to MLAG from PIM",
__func__);
router->connected_to_mlag = false;
zclient_send_mlag_deregister(zclient);
return 0;
}
void pim_mlag_deregister(void)
{
/* if somebody still interested in the MLAG channel skip de-reg */
if (router->pim_mlag_intf_cnt || pim_vxlan_do_mlag_reg())
return;
/* not registered; nothing do */
if (!router->mlag_process_register)
return;
router->mlag_process_register = false;
thread_add_event(router->master, pim_mlag_deregister_handler, NULL, 0,
NULL);
}
void pim_if_configure_mlag_dualactive(struct pim_interface *pim_ifp)
{