forked from openvswitch/ovs
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathofproto-dpif-sflow.c
1407 lines (1262 loc) · 45 KB
/
ofproto-dpif-sflow.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
/*
* Copyright (c) 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016 Nicira, Inc.
* Copyright (c) 2009 InMon Corp.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at:
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include <config.h>
#include "ofproto-dpif-sflow.h"
#include <inttypes.h>
#include <sys/resource.h>
#include <sys/socket.h>
#include <net/if.h>
#include <stdlib.h>
#include "collectors.h"
#include "compiler.h"
#include "dpif.h"
#include "hash.h"
#include "openvswitch/hmap.h"
#include "netdev.h"
#include "netlink.h"
#include "openvswitch/ofpbuf.h"
#include "ofproto.h"
#include "packets.h"
#include "poll-loop.h"
#include "ovs-router.h"
#include "route-table.h"
#include "sflow_api.h"
#include "socket-util.h"
#include "timeval.h"
#include "openvswitch/vlog.h"
#include "lib/odp-util.h"
#include "lib/unaligned.h"
#include "ofproto-provider.h"
#include "lacp.h"
VLOG_DEFINE_THIS_MODULE(sflow);
static struct ovs_mutex mutex;
/* This global var is used to determine which sFlow
sub-agent should send the datapath counters. */
#define SFLOW_GC_SUBID_UNCLAIMED (uint32_t)-1
static uint32_t sflow_global_counters_subid = SFLOW_GC_SUBID_UNCLAIMED;
/*
* The enum dpif_sflow_tunnel_type is to declare the types supported
*/
enum dpif_sflow_tunnel_type {
DPIF_SFLOW_TUNNEL_UNKNOWN = 0,
DPIF_SFLOW_TUNNEL_VXLAN,
DPIF_SFLOW_TUNNEL_GRE,
DPIF_SFLOW_TUNNEL_LISP,
DPIF_SFLOW_TUNNEL_GENEVE
};
struct dpif_sflow_port {
struct hmap_node hmap_node; /* In struct dpif_sflow's "ports" hmap. */
SFLDataSource_instance dsi; /* sFlow library's notion of port number. */
struct ofport *ofport; /* To retrive port stats. */
odp_port_t odp_port;
enum dpif_sflow_tunnel_type tunnel_type;
};
struct dpif_sflow {
struct collectors *collectors;
SFLAgent *sflow_agent;
struct ofproto_sflow_options *options;
time_t next_tick;
size_t n_flood, n_all;
struct hmap ports; /* Contains "struct dpif_sflow_port"s. */
uint32_t probability;
struct ovs_refcount ref_cnt;
};
static void dpif_sflow_del_port__(struct dpif_sflow *,
struct dpif_sflow_port *);
#define RECEIVER_INDEX 1
static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 5);
static bool
ofproto_sflow_options_equal(const struct ofproto_sflow_options *a,
const struct ofproto_sflow_options *b)
{
return (sset_equals(&a->targets, &b->targets)
&& a->sampling_rate == b->sampling_rate
&& a->polling_interval == b->polling_interval
&& a->header_len == b->header_len
&& a->sub_id == b->sub_id
&& nullable_string_is_equal(a->agent_device, b->agent_device)
&& nullable_string_is_equal(a->control_ip, b->control_ip));
}
static struct ofproto_sflow_options *
ofproto_sflow_options_clone(const struct ofproto_sflow_options *old)
{
struct ofproto_sflow_options *new = xmemdup(old, sizeof *old);
sset_clone(&new->targets, &old->targets);
new->agent_device = nullable_xstrdup(old->agent_device);
new->control_ip = nullable_xstrdup(old->control_ip);
return new;
}
static void
ofproto_sflow_options_destroy(struct ofproto_sflow_options *options)
{
if (options) {
sset_destroy(&options->targets);
free(options->agent_device);
free(options->control_ip);
free(options);
}
}
/* sFlow library callback to allocate memory. */
static void *
sflow_agent_alloc_cb(void *magic OVS_UNUSED, SFLAgent *agent OVS_UNUSED,
size_t bytes)
{
return xzalloc(bytes);
}
/* sFlow library callback to free memory. */
static int
sflow_agent_free_cb(void *magic OVS_UNUSED, SFLAgent *agent OVS_UNUSED,
void *obj)
{
free(obj);
return 0;
}
/* sFlow library callback to report error. */
static void
sflow_agent_error_cb(void *magic OVS_UNUSED, SFLAgent *agent OVS_UNUSED,
char *msg)
{
VLOG_WARN("sFlow agent error: %s", msg);
}
/* sFlow library callback to send datagram. */
static void
sflow_agent_send_packet_cb(void *ds_, SFLAgent *agent OVS_UNUSED,
SFLReceiver *receiver OVS_UNUSED, u_char *pkt,
uint32_t pktLen)
{
struct dpif_sflow *ds = ds_;
collectors_send(ds->collectors, pkt, pktLen);
}
static struct dpif_sflow_port *
dpif_sflow_find_port(const struct dpif_sflow *ds, odp_port_t odp_port)
OVS_REQUIRES(mutex)
{
struct dpif_sflow_port *dsp;
HMAP_FOR_EACH_IN_BUCKET (dsp, hmap_node, hash_odp_port(odp_port),
&ds->ports) {
if (dsp->odp_port == odp_port) {
return dsp;
}
}
return NULL;
}
/* Call to get the datapath stats. Modeled after the dpctl utility.
*
* It might be more efficient for this module to be given a handle it can use
* to get these stats more efficiently, but this is only going to be called
* once every 20-30 seconds. Return number of datapaths found (normally expect
* 1). */
static int
sflow_get_dp_stats(struct dpif_sflow *ds OVS_UNUSED,
struct dpif_dp_stats *dp_totals)
{
struct sset types;
const char *type;
int count = 0;
memset(dp_totals, 0, sizeof *dp_totals);
sset_init(&types);
dp_enumerate_types(&types);
SSET_FOR_EACH (type, &types) {
struct sset names;
const char *name;
sset_init(&names);
if (dp_enumerate_names(type, &names) == 0) {
SSET_FOR_EACH (name, &names) {
struct dpif *dpif;
if (dpif_open(name, type, &dpif) == 0) {
struct dpif_dp_stats dp_stats;
if (dpif_get_dp_stats(dpif, &dp_stats) == 0) {
count++;
dp_totals->n_hit += dp_stats.n_hit;
dp_totals->n_missed += dp_stats.n_missed;
dp_totals->n_lost += dp_stats.n_lost;
dp_totals->n_flows += dp_stats.n_flows;
dp_totals->n_mask_hit += dp_stats.n_mask_hit;
dp_totals->n_masks += dp_stats.n_masks;
}
dpif_close(dpif);
}
}
sset_destroy(&names);
}
}
sset_destroy(&types);
return count;
}
/* If there are multiple bridges defined then we need some
minimal artibration to decide which one should send the
global counters. This function allows each sub-agent to
ask if he should do it or not. */
static bool
sflow_global_counters_subid_test(uint32_t subid)
OVS_REQUIRES(mutex)
{
if (sflow_global_counters_subid == SFLOW_GC_SUBID_UNCLAIMED) {
/* The role is up for grabs. */
sflow_global_counters_subid = subid;
}
return (sflow_global_counters_subid == subid);
}
static void
sflow_global_counters_subid_clear(uint32_t subid)
OVS_REQUIRES(mutex)
{
if (sflow_global_counters_subid == subid) {
/* The sub-agent that was sending global counters
is going away, so reset to allow another
to take over. */
sflow_global_counters_subid = SFLOW_GC_SUBID_UNCLAIMED;
}
}
static void
sflow_agent_get_global_counters(void *ds_, SFLPoller *poller,
SFL_COUNTERS_SAMPLE_TYPE *cs)
OVS_REQUIRES(mutex)
{
struct dpif_sflow *ds = ds_;
SFLCounters_sample_element dp_elem, res_elem;
struct dpif_dp_stats dp_totals;
struct rusage usage;
if (!sflow_global_counters_subid_test(poller->agent->subId)) {
/* Another sub-agent is currently responsible for this. */
return;
}
/* datapath stats */
if (sflow_get_dp_stats(ds, &dp_totals)) {
dp_elem.tag = SFLCOUNTERS_OVSDP;
dp_elem.counterBlock.ovsdp.n_hit = dp_totals.n_hit;
dp_elem.counterBlock.ovsdp.n_missed = dp_totals.n_missed;
dp_elem.counterBlock.ovsdp.n_lost = dp_totals.n_lost;
dp_elem.counterBlock.ovsdp.n_mask_hit = dp_totals.n_mask_hit;
dp_elem.counterBlock.ovsdp.n_flows = dp_totals.n_flows;
dp_elem.counterBlock.ovsdp.n_masks = dp_totals.n_masks;
SFLADD_ELEMENT(cs, &dp_elem);
}
/* resource usage */
getrusage(RUSAGE_SELF, &usage);
res_elem.tag = SFLCOUNTERS_APP_RESOURCES;
res_elem.counterBlock.appResources.user_time
= timeval_to_msec(&usage.ru_utime);
res_elem.counterBlock.appResources.system_time
= timeval_to_msec(&usage.ru_stime);
res_elem.counterBlock.appResources.mem_used = (usage.ru_maxrss * 1024);
SFL_UNDEF_GAUGE(res_elem.counterBlock.appResources.mem_max);
SFL_UNDEF_GAUGE(res_elem.counterBlock.appResources.fd_open);
SFL_UNDEF_GAUGE(res_elem.counterBlock.appResources.fd_max);
SFL_UNDEF_GAUGE(res_elem.counterBlock.appResources.conn_open);
SFL_UNDEF_GAUGE(res_elem.counterBlock.appResources.conn_max);
SFLADD_ELEMENT(cs, &res_elem);
sfl_poller_writeCountersSample(poller, cs);
}
static void
sflow_agent_get_counters(void *ds_, SFLPoller *poller,
SFL_COUNTERS_SAMPLE_TYPE *cs)
OVS_REQUIRES(mutex)
{
struct dpif_sflow *ds = ds_;
SFLCounters_sample_element elem, lacp_elem, of_elem, name_elem;
SFLCounters_sample_element eth_elem;
enum netdev_features current;
struct dpif_sflow_port *dsp;
SFLIf_counters *counters;
SFLEthernet_counters* eth_counters;
struct netdev_stats stats;
enum netdev_flags flags;
struct lacp_slave_stats lacp_stats;
const char *ifName;
dsp = dpif_sflow_find_port(ds, u32_to_odp(poller->bridgePort));
if (!dsp) {
return;
}
elem.tag = SFLCOUNTERS_GENERIC;
counters = &elem.counterBlock.generic;
counters->ifIndex = SFL_DS_INDEX(poller->dsi);
counters->ifType = 6;
if (!netdev_get_features(dsp->ofport->netdev, ¤t, NULL, NULL, NULL)) {
/* The values of ifDirection come from MAU MIB (RFC 2668): 0 = unknown,
1 = full-duplex, 2 = half-duplex, 3 = in, 4=out */
counters->ifSpeed = netdev_features_to_bps(current, 0);
counters->ifDirection = (netdev_features_is_full_duplex(current)
? 1 : 2);
} else {
counters->ifSpeed = 100000000;
counters->ifDirection = 0;
}
if (!netdev_get_flags(dsp->ofport->netdev, &flags) && flags & NETDEV_UP) {
counters->ifStatus = 1; /* ifAdminStatus up. */
if (netdev_get_carrier(dsp->ofport->netdev)) {
counters->ifStatus |= 2; /* ifOperStatus us. */
}
} else {
counters->ifStatus = 0; /* Down. */
}
/* XXX
1. Is the multicast counter filled in?
2. Does the multicast counter include broadcasts?
3. Does the rx_packets counter include multicasts/broadcasts?
*/
ofproto_port_get_stats(dsp->ofport, &stats);
counters->ifInOctets = stats.rx_bytes;
counters->ifInUcastPkts = stats.rx_packets;
counters->ifInMulticastPkts = stats.multicast;
counters->ifInBroadcastPkts = stats.rx_broadcast_packets;
counters->ifInDiscards = stats.rx_dropped;
counters->ifInErrors = stats.rx_errors;
counters->ifInUnknownProtos = -1;
counters->ifOutOctets = stats.tx_bytes;
counters->ifOutUcastPkts = stats.tx_packets;
counters->ifOutMulticastPkts = stats.tx_multicast_packets;
counters->ifOutBroadcastPkts = stats.tx_broadcast_packets;
counters->ifOutDiscards = stats.tx_dropped;
counters->ifOutErrors = stats.tx_errors;
counters->ifPromiscuousMode = 0;
SFLADD_ELEMENT(cs, &elem);
/* Include LACP counters and identifiers if this port is part of a LAG. */
if (ofproto_port_get_lacp_stats(dsp->ofport, &lacp_stats) == 0) {
memset(&lacp_elem, 0, sizeof lacp_elem);
lacp_elem.tag = SFLCOUNTERS_LACP;
lacp_elem.counterBlock.lacp.actorSystemID =
lacp_stats.dot3adAggPortActorSystemID;
lacp_elem.counterBlock.lacp.partnerSystemID =
lacp_stats.dot3adAggPortPartnerOperSystemID;
lacp_elem.counterBlock.lacp.attachedAggID =
lacp_stats.dot3adAggPortAttachedAggID;
lacp_elem.counterBlock.lacp.portState.v.actorAdmin =
lacp_stats.dot3adAggPortActorAdminState;
lacp_elem.counterBlock.lacp.portState.v.actorOper =
lacp_stats.dot3adAggPortActorOperState;
lacp_elem.counterBlock.lacp.portState.v.partnerAdmin =
lacp_stats.dot3adAggPortPartnerAdminState;
lacp_elem.counterBlock.lacp.portState.v.partnerOper =
lacp_stats.dot3adAggPortPartnerOperState;
lacp_elem.counterBlock.lacp.LACPDUsRx =
lacp_stats.dot3adAggPortStatsLACPDUsRx;
SFL_UNDEF_COUNTER(lacp_elem.counterBlock.lacp.markerPDUsRx);
SFL_UNDEF_COUNTER(lacp_elem.counterBlock.lacp.markerResponsePDUsRx);
SFL_UNDEF_COUNTER(lacp_elem.counterBlock.lacp.unknownRx);
lacp_elem.counterBlock.lacp.illegalRx =
lacp_stats.dot3adAggPortStatsIllegalRx;
lacp_elem.counterBlock.lacp.LACPDUsTx =
lacp_stats.dot3adAggPortStatsLACPDUsTx;
SFL_UNDEF_COUNTER(lacp_elem.counterBlock.lacp.markerPDUsTx);
SFL_UNDEF_COUNTER(lacp_elem.counterBlock.lacp.markerResponsePDUsTx);
SFLADD_ELEMENT(cs, &lacp_elem);
}
/* Include Port name. */
if ((ifName = netdev_get_name(dsp->ofport->netdev)) != NULL) {
memset(&name_elem, 0, sizeof name_elem);
name_elem.tag = SFLCOUNTERS_PORTNAME;
name_elem.counterBlock.portName.portName.str = (char *)ifName;
name_elem.counterBlock.portName.portName.len = strlen(ifName);
SFLADD_ELEMENT(cs, &name_elem);
}
/* Include OpenFlow DPID and openflow port number. */
memset(&of_elem, 0, sizeof of_elem);
of_elem.tag = SFLCOUNTERS_OPENFLOWPORT;
of_elem.counterBlock.ofPort.datapath_id =
ofproto_get_datapath_id(dsp->ofport->ofproto);
of_elem.counterBlock.ofPort.port_no =
(OVS_FORCE uint32_t)dsp->ofport->ofp_port;
SFLADD_ELEMENT(cs, &of_elem);
/* Include ethernet counters */
memset(ð_elem, 0, sizeof eth_elem);
eth_elem.tag = SFLCOUNTERS_ETHERNET;
eth_counters = ð_elem.counterBlock.ethernet;
eth_counters->dot3StatsAlignmentErrors = stats.rx_frame_errors;
eth_counters->dot3StatsFCSErrors = stats.rx_crc_errors;
eth_counters->dot3StatsFrameTooLongs = stats.rx_oversize_errors;
SFL_UNDEF_COUNTER(eth_counters->dot3StatsSingleCollisionFrames);
SFL_UNDEF_COUNTER(eth_counters->dot3StatsMultipleCollisionFrames);
SFL_UNDEF_COUNTER(eth_counters->dot3StatsSQETestErrors);
SFL_UNDEF_COUNTER(eth_counters->dot3StatsDeferredTransmissions);
SFL_UNDEF_COUNTER(eth_counters->dot3StatsLateCollisions);
SFL_UNDEF_COUNTER(eth_counters->dot3StatsExcessiveCollisions);
SFL_UNDEF_COUNTER(eth_counters->dot3StatsInternalMacTransmitErrors);
SFL_UNDEF_COUNTER(eth_counters->dot3StatsCarrierSenseErrors);
SFL_UNDEF_COUNTER(eth_counters->dot3StatsInternalMacReceiveErrors);
SFL_UNDEF_COUNTER(eth_counters->dot3StatsSymbolErrors);
SFLADD_ELEMENT(cs, ð_elem);
sfl_poller_writeCountersSample(poller, cs);
}
/* Obtains an address to use for the local sFlow agent and stores it into
* '*agent_addr'. Returns true if successful, false on failure.
*
* The sFlow agent address should be a local IP address that is persistent and
* reachable over the network, if possible. The IP address associated with
* 'agent_device' is used if it has one, and otherwise 'control_ip', the IP
* address used to talk to the controller. If the agent device is not
* specified then it is figured out by taking a look at the routing table based
* on 'targets'. */
static bool
sflow_choose_agent_address(const char *agent_device,
const struct sset *targets,
const char *control_ip,
SFLAddress *agent_addr)
{
const char *target;
struct in_addr in4;
memset(agent_addr, 0, sizeof *agent_addr);
agent_addr->type = SFLADDRESSTYPE_IP_V4;
if (agent_device) {
if (!netdev_get_in4_by_name(agent_device, &in4)) {
goto success;
}
}
SSET_FOR_EACH (target, targets) {
union {
struct sockaddr_storage ss;
struct sockaddr_in sin;
} sa;
char name[IFNAMSIZ];
if (inet_parse_active(target, SFL_DEFAULT_COLLECTOR_PORT, &sa.ss)
&& sa.ss.ss_family == AF_INET) {
struct in6_addr addr6, src, gw;
in6_addr_set_mapped_ipv4(&addr6, sa.sin.sin_addr.s_addr);
/* sFlow only supports target in default routing table with
* packet mark zero.
*/
if (ovs_router_lookup(0, &addr6, name, &src, &gw)) {
in4.s_addr = in6_addr_get_mapped_ipv4(&src);
goto success;
}
}
}
if (control_ip && !lookup_ip(control_ip, &in4)) {
goto success;
}
VLOG_ERR("could not determine IP address for sFlow agent");
return false;
success:
agent_addr->address.ip_v4.addr = (OVS_FORCE uint32_t) in4.s_addr;
return true;
}
static void
dpif_sflow_clear__(struct dpif_sflow *ds) OVS_REQUIRES(mutex)
{
if (ds->sflow_agent) {
sflow_global_counters_subid_clear(ds->sflow_agent->subId);
sfl_agent_release(ds->sflow_agent);
free(ds->sflow_agent);
ds->sflow_agent = NULL;
}
collectors_destroy(ds->collectors);
ds->collectors = NULL;
ofproto_sflow_options_destroy(ds->options);
ds->options = NULL;
/* Turn off sampling to save CPU cycles. */
ds->probability = 0;
}
void
dpif_sflow_clear(struct dpif_sflow *ds) OVS_EXCLUDED(mutex)
{
ovs_mutex_lock(&mutex);
dpif_sflow_clear__(ds);
ovs_mutex_unlock(&mutex);
}
bool
dpif_sflow_is_enabled(const struct dpif_sflow *ds) OVS_EXCLUDED(mutex)
{
bool enabled;
ovs_mutex_lock(&mutex);
enabled = ds->collectors != NULL;
ovs_mutex_unlock(&mutex);
return enabled;
}
struct dpif_sflow *
dpif_sflow_create(void)
{
static struct ovsthread_once once = OVSTHREAD_ONCE_INITIALIZER;
struct dpif_sflow *ds;
if (ovsthread_once_start(&once)) {
ovs_mutex_init_recursive(&mutex);
ovsthread_once_done(&once);
}
ds = xcalloc(1, sizeof *ds);
ds->next_tick = time_now() + 1;
hmap_init(&ds->ports);
ds->probability = 0;
ovs_refcount_init(&ds->ref_cnt);
return ds;
}
struct dpif_sflow *
dpif_sflow_ref(const struct dpif_sflow *ds_)
{
struct dpif_sflow *ds = CONST_CAST(struct dpif_sflow *, ds_);
if (ds) {
ovs_refcount_ref(&ds->ref_cnt);
}
return ds;
}
/* 32-bit fraction of packets to sample with. A value of 0 samples no packets,
* a value of %UINT32_MAX samples all packets and intermediate values sample
* intermediate fractions of packets. */
uint32_t
dpif_sflow_get_probability(const struct dpif_sflow *ds) OVS_EXCLUDED(mutex)
{
uint32_t probability;
ovs_mutex_lock(&mutex);
probability = ds->probability;
ovs_mutex_unlock(&mutex);
return probability;
}
void
dpif_sflow_unref(struct dpif_sflow *ds) OVS_EXCLUDED(mutex)
{
if (ds && ovs_refcount_unref_relaxed(&ds->ref_cnt) == 1) {
struct dpif_sflow_port *dsp, *next;
dpif_sflow_clear(ds);
HMAP_FOR_EACH_SAFE (dsp, next, hmap_node, &ds->ports) {
dpif_sflow_del_port__(ds, dsp);
}
hmap_destroy(&ds->ports);
free(ds);
}
}
static void
dpif_sflow_add_poller(struct dpif_sflow *ds, struct dpif_sflow_port *dsp)
OVS_REQUIRES(mutex)
{
SFLPoller *poller = sfl_agent_addPoller(ds->sflow_agent, &dsp->dsi, ds,
sflow_agent_get_counters);
sfl_poller_set_sFlowCpInterval(poller, ds->options->polling_interval);
sfl_poller_set_sFlowCpReceiver(poller, RECEIVER_INDEX);
sfl_poller_set_bridgePort(poller, odp_to_u32(dsp->odp_port));
}
static enum dpif_sflow_tunnel_type
dpif_sflow_tunnel_type(struct ofport *ofport) {
const char *type = netdev_get_type(ofport->netdev);
if (type) {
if (strcmp(type, "gre") == 0) {
return DPIF_SFLOW_TUNNEL_GRE;
} else if (strcmp(type, "vxlan") == 0) {
return DPIF_SFLOW_TUNNEL_VXLAN;
} else if (strcmp(type, "lisp") == 0) {
return DPIF_SFLOW_TUNNEL_LISP;
} else if (strcmp(type, "geneve") == 0) {
return DPIF_SFLOW_TUNNEL_GENEVE;
}
}
return DPIF_SFLOW_TUNNEL_UNKNOWN;
}
static uint8_t
dpif_sflow_tunnel_proto(enum dpif_sflow_tunnel_type tunnel_type)
{
/* Default to 0 (IPPROTO_IP), meaning "unknown". */
uint8_t ipproto = 0;
switch(tunnel_type) {
case DPIF_SFLOW_TUNNEL_GRE:
ipproto = IPPROTO_GRE;
break;
case DPIF_SFLOW_TUNNEL_VXLAN:
case DPIF_SFLOW_TUNNEL_LISP:
case DPIF_SFLOW_TUNNEL_GENEVE:
ipproto = IPPROTO_UDP;
case DPIF_SFLOW_TUNNEL_UNKNOWN:
break;
}
return ipproto;
}
void
dpif_sflow_add_port(struct dpif_sflow *ds, struct ofport *ofport,
odp_port_t odp_port) OVS_EXCLUDED(mutex)
{
struct dpif_sflow_port *dsp;
int ifindex;
enum dpif_sflow_tunnel_type tunnel_type;
ovs_mutex_lock(&mutex);
dpif_sflow_del_port(ds, odp_port);
tunnel_type = dpif_sflow_tunnel_type(ofport);
ifindex = netdev_get_ifindex(ofport->netdev);
if (ifindex <= 0
&& tunnel_type == DPIF_SFLOW_TUNNEL_UNKNOWN) {
/* Not an ifindex port, and not a tunnel port either
* so do not add a cross-reference to it here.
*/
goto out;
}
/* Add to table of ports. */
dsp = xmalloc(sizeof *dsp);
dsp->ofport = ofport;
dsp->odp_port = odp_port;
dsp->tunnel_type = tunnel_type;
hmap_insert(&ds->ports, &dsp->hmap_node, hash_odp_port(odp_port));
if (ifindex > 0) {
/* Add poller for ports that have ifindex. */
SFL_DS_SET(dsp->dsi, SFL_DSCLASS_IFINDEX, ifindex, 0);
if (ds->sflow_agent) {
dpif_sflow_add_poller(ds, dsp);
}
} else {
/* Record "ifindex unknown" for the others */
SFL_DS_SET(dsp->dsi, SFL_DSCLASS_IFINDEX, 0, 0);
}
out:
ovs_mutex_unlock(&mutex);
}
static void
dpif_sflow_del_port__(struct dpif_sflow *ds, struct dpif_sflow_port *dsp)
OVS_REQUIRES(mutex)
{
if (ds->sflow_agent
&& SFL_DS_INDEX(dsp->dsi)) {
sfl_agent_removePoller(ds->sflow_agent, &dsp->dsi);
sfl_agent_removeSampler(ds->sflow_agent, &dsp->dsi);
}
hmap_remove(&ds->ports, &dsp->hmap_node);
free(dsp);
}
void
dpif_sflow_del_port(struct dpif_sflow *ds, odp_port_t odp_port)
OVS_EXCLUDED(mutex)
{
struct dpif_sflow_port *dsp;
ovs_mutex_lock(&mutex);
dsp = dpif_sflow_find_port(ds, odp_port);
if (dsp) {
dpif_sflow_del_port__(ds, dsp);
}
ovs_mutex_unlock(&mutex);
}
void
dpif_sflow_set_options(struct dpif_sflow *ds,
const struct ofproto_sflow_options *options)
OVS_EXCLUDED(mutex)
{
struct dpif_sflow_port *dsp;
bool options_changed;
SFLReceiver *receiver;
SFLAddress agentIP;
time_t now;
SFLDataSource_instance dsi;
uint32_t dsIndex;
SFLSampler *sampler;
SFLPoller *poller;
ovs_mutex_lock(&mutex);
if (sset_is_empty(&options->targets) || !options->sampling_rate) {
/* No point in doing any work if there are no targets or nothing to
* sample. */
dpif_sflow_clear__(ds);
goto out;
}
options_changed = (!ds->options
|| !ofproto_sflow_options_equal(options, ds->options));
/* Configure collectors if options have changed or if we're shortchanged in
* collectors (which indicates that opening one or more of the configured
* collectors failed, so that we should retry). */
if (options_changed
|| collectors_count(ds->collectors) < sset_count(&options->targets)) {
collectors_destroy(ds->collectors);
collectors_create(&options->targets, SFL_DEFAULT_COLLECTOR_PORT,
&ds->collectors);
if (ds->collectors == NULL) {
VLOG_WARN_RL(&rl, "no collectors could be initialized, "
"sFlow disabled");
dpif_sflow_clear__(ds);
goto out;
}
}
/* Choose agent IP address and agent device (if not yet setup) */
if (!sflow_choose_agent_address(options->agent_device,
&options->targets,
options->control_ip, &agentIP)) {
dpif_sflow_clear__(ds);
goto out;
}
/* Avoid reconfiguring if options didn't change. */
if (!options_changed) {
goto out;
}
ofproto_sflow_options_destroy(ds->options);
ds->options = ofproto_sflow_options_clone(options);
/* Create agent. */
VLOG_INFO("creating sFlow agent %d", options->sub_id);
if (ds->sflow_agent) {
sflow_global_counters_subid_clear(ds->sflow_agent->subId);
sfl_agent_release(ds->sflow_agent);
}
ds->sflow_agent = xcalloc(1, sizeof *ds->sflow_agent);
now = time_wall();
sfl_agent_init(ds->sflow_agent,
&agentIP,
options->sub_id,
now, /* Boot time. */
now, /* Current time. */
ds, /* Pointer supplied to callbacks. */
sflow_agent_alloc_cb,
sflow_agent_free_cb,
sflow_agent_error_cb,
sflow_agent_send_packet_cb);
receiver = sfl_agent_addReceiver(ds->sflow_agent);
sfl_receiver_set_sFlowRcvrOwner(receiver, "Open vSwitch sFlow");
sfl_receiver_set_sFlowRcvrTimeout(receiver, 0xffffffff);
/* Set the sampling_rate down in the datapath. */
ds->probability = MAX(1, UINT32_MAX / ds->options->sampling_rate);
/* Add a single sampler for the bridge. This appears as a PHYSICAL_ENTITY
because it is associated with the hypervisor, and interacts with the server
hardware directly. The sub_id is used to distinguish this sampler from
others on other bridges within the same agent. */
dsIndex = 1000 + options->sub_id;
SFL_DS_SET(dsi, SFL_DSCLASS_PHYSICAL_ENTITY, dsIndex, 0);
sampler = sfl_agent_addSampler(ds->sflow_agent, &dsi);
sfl_sampler_set_sFlowFsPacketSamplingRate(sampler, ds->options->sampling_rate);
sfl_sampler_set_sFlowFsMaximumHeaderSize(sampler, ds->options->header_len);
sfl_sampler_set_sFlowFsReceiver(sampler, RECEIVER_INDEX);
/* Add a counter poller for the bridge so we can use it to send
global counters such as datapath cache hit/miss stats. */
poller = sfl_agent_addPoller(ds->sflow_agent, &dsi, ds,
sflow_agent_get_global_counters);
sfl_poller_set_sFlowCpInterval(poller, ds->options->polling_interval);
sfl_poller_set_sFlowCpReceiver(poller, RECEIVER_INDEX);
/* Add pollers for the currently known ifindex-ports */
HMAP_FOR_EACH (dsp, hmap_node, &ds->ports) {
if (SFL_DS_INDEX(dsp->dsi)) {
dpif_sflow_add_poller(ds, dsp);
}
}
out:
ovs_mutex_unlock(&mutex);
}
int
dpif_sflow_odp_port_to_ifindex(const struct dpif_sflow *ds,
odp_port_t odp_port) OVS_EXCLUDED(mutex)
{
struct dpif_sflow_port *dsp;
int ret;
ovs_mutex_lock(&mutex);
dsp = dpif_sflow_find_port(ds, odp_port);
ret = dsp ? SFL_DS_INDEX(dsp->dsi) : 0;
ovs_mutex_unlock(&mutex);
return ret;
}
static void
dpif_sflow_tunnel_v4(uint8_t tunnel_ipproto,
const struct flow_tnl *tunnel,
SFLSampled_ipv4 *ipv4)
{
ipv4->protocol = tunnel_ipproto;
ipv4->tos = tunnel->ip_tos;
ipv4->src_ip.addr = (OVS_FORCE uint32_t) tunnel->ip_src;
ipv4->dst_ip.addr = (OVS_FORCE uint32_t) tunnel->ip_dst;
ipv4->src_port = (OVS_FORCE uint16_t) tunnel->tp_src;
ipv4->dst_port = (OVS_FORCE uint16_t) tunnel->tp_dst;
}
static void
dpif_sflow_push_mpls_lse(struct dpif_sflow_actions *sflow_actions,
ovs_be32 lse)
{
if (sflow_actions->mpls_stack_depth >= FLOW_MAX_MPLS_LABELS) {
sflow_actions->mpls_err = true;
return;
}
/* Record the new lse in host-byte-order. */
/* BOS flag will be fixed later when we send stack to sFlow library. */
sflow_actions->mpls_lse[sflow_actions->mpls_stack_depth++] = ntohl(lse);
}
static void
dpif_sflow_pop_mpls_lse(struct dpif_sflow_actions *sflow_actions)
{
if (sflow_actions->mpls_stack_depth == 0) {
sflow_actions->mpls_err = true;
return;
}
sflow_actions->mpls_stack_depth--;
}
static void
dpif_sflow_set_mpls(struct dpif_sflow_actions *sflow_actions,
const struct ovs_key_mpls *mpls_key, int n)
{
int ii;
if (n > FLOW_MAX_MPLS_LABELS) {
sflow_actions->mpls_err = true;
return;
}
for (ii = 0; ii < n; ii++) {
/* Reverse stack order, and use host-byte-order for each lse. */
sflow_actions->mpls_lse[n - ii - 1] = ntohl(mpls_key[ii].mpls_lse);
}
sflow_actions->mpls_stack_depth = n;
}
static void
sflow_read_tnl_push_action(const struct nlattr *attr,
struct dpif_sflow_actions *sflow_actions)
{
/* Modeled on lib/odp-util.c: format_odp_tnl_push_header */
const struct ovs_action_push_tnl *data = nl_attr_get(attr);
const struct eth_header *eth = (const struct eth_header *) data->header;
const struct ip_header *ip
= ALIGNED_CAST(const struct ip_header *, eth + 1);
sflow_actions->out_port = u32_to_odp(data->out_port);
/* Ethernet. */
/* TODO: SFlow does not currently define a MAC-in-MAC
* encapsulation structure. We could use an extension
* structure to report this.
*/
/* IPv4 */
/* Cannot assume alignment so just use memcpy. */
sflow_actions->tunnel.ip_src = get_16aligned_be32(&ip->ip_src);
sflow_actions->tunnel.ip_dst = get_16aligned_be32(&ip->ip_dst);
sflow_actions->tunnel.ip_tos = ip->ip_tos;
sflow_actions->tunnel.ip_ttl = ip->ip_ttl;
/* The tnl_push action can supply the ip_protocol too. */
sflow_actions->tunnel_ipproto = ip->ip_proto;
/* Layer 4 */
if (data->tnl_type == OVS_VPORT_TYPE_VXLAN
|| data->tnl_type == OVS_VPORT_TYPE_GENEVE) {
const struct udp_header *udp = (const struct udp_header *) (ip + 1);
sflow_actions->tunnel.tp_src = udp->udp_src;
sflow_actions->tunnel.tp_dst = udp->udp_dst;
if (data->tnl_type == OVS_VPORT_TYPE_VXLAN) {
const struct vxlanhdr *vxh = (const struct vxlanhdr *) (udp + 1);
uint64_t tun_id = ntohl(get_16aligned_be32(&vxh->vx_vni)) >> 8;
sflow_actions->tunnel.tun_id = htonll(tun_id);
} else {
const struct genevehdr *gnh = (const struct genevehdr *) (udp + 1);
uint64_t tun_id = ntohl(get_16aligned_be32(&gnh->vni)) >> 8;
sflow_actions->tunnel.tun_id = htonll(tun_id);
}
} else if (data->tnl_type == OVS_VPORT_TYPE_GRE) {
const void *l4 = ip + 1;
const struct gre_base_hdr *greh = (const struct gre_base_hdr *) l4;
ovs_16aligned_be32 *options = (ovs_16aligned_be32 *)(greh + 1);
if (greh->flags & htons(GRE_CSUM)) {
options++;
}
if (greh->flags & htons(GRE_KEY)) {
uint64_t tun_id = ntohl(get_16aligned_be32(options));
sflow_actions->tunnel.tun_id = htonll(tun_id);
}
}
}
static void
sflow_read_set_action(const struct nlattr *attr,
struct dpif_sflow_actions *sflow_actions)
{
enum ovs_key_attr type = nl_attr_type(attr);
switch (type) {
case OVS_KEY_ATTR_ENCAP:
if (++sflow_actions->encap_depth > 1) {
/* Do not handle multi-encap for now. */
sflow_actions->tunnel_err = true;
} else {
dpif_sflow_read_actions(NULL,
nl_attr_get(attr), nl_attr_get_size(attr),
sflow_actions);
}
break;
case OVS_KEY_ATTR_PRIORITY:
case OVS_KEY_ATTR_SKB_MARK:
case OVS_KEY_ATTR_DP_HASH:
case OVS_KEY_ATTR_RECIRC_ID:
break;
case OVS_KEY_ATTR_TUNNEL: {
if (++sflow_actions->encap_depth > 1) {
/* Do not handle multi-encap for now. */
sflow_actions->tunnel_err = true;
} else {
if (odp_tun_key_from_attr(attr, &sflow_actions->tunnel)
== ODP_FIT_ERROR) {
/* Tunnel parsing error. */
sflow_actions->tunnel_err = true;
}
}
break;
}
case OVS_KEY_ATTR_IN_PORT:
case OVS_KEY_ATTR_ETHERNET:
case OVS_KEY_ATTR_VLAN:
break;
case OVS_KEY_ATTR_MPLS: {
const struct ovs_key_mpls *mpls_key = nl_attr_get(attr);
size_t size = nl_attr_get_size(attr);
dpif_sflow_set_mpls(sflow_actions, mpls_key, size / sizeof *mpls_key);
break;
}
case OVS_KEY_ATTR_ETHERTYPE:
case OVS_KEY_ATTR_IPV4: