summaryrefslogtreecommitdiffstats
path: root/xlators/mgmt/glusterd/src/glusterd-syncop.c
blob: 7b0c28baf1aac5580973c1565de95eae2510694c (plain)
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
/*
   Copyright (c) 2012-2012 Red Hat, Inc. <http://www.redhat.com>
   This file is part of GlusterFS.

   This file is licensed to you under your choice of the GNU Lesser
   General Public License, version 3 or any later version (LGPLv3 or
   later), or the GNU General Public License, version 2 (GPLv2), in all
   cases as published by the Free Software Foundation.
*/
/* rpc related syncops */
#include "rpc-clnt.h"
#include "protocol-common.h"
#include "xdr-generic.h"
#include "glusterd1-xdr.h"
#include "glusterd-syncop.h"

#include "glusterd.h"
#include "glusterd-op-sm.h"
#include "glusterd-utils.h"

static inline void
gd_synctask_barrier_wait (struct syncargs *args, int count)
{
        glusterd_conf_t *conf = THIS->private;

        synclock_unlock (&conf->big_lock);
        synctask_barrier_wait (args, count);
        synclock_lock (&conf->big_lock);

	syncbarrier_destroy (&args->barrier);
}

static void
gd_collate_errors (struct syncargs *args, int op_ret, int op_errno,
                   char *op_errstr)
{
        if (args->op_ret)
                return;
        args->op_ret = op_ret;
        args->op_errno = op_errno;
        if (op_ret && op_errstr && strcmp (op_errstr, ""))
                args->errstr = gf_strdup (op_errstr);
}

static void
gd_syncargs_init (struct syncargs *args, dict_t *op_ctx)
{
        args->dict = op_ctx;
        pthread_mutex_init (&args->lock_dict, NULL);
}

static void
gd_stage_op_req_free (gd1_mgmt_stage_op_req *req)
{
        if (!req)
                return;

        GF_FREE (req->buf.buf_val);
        GF_FREE (req);
}

static void
gd_commit_op_req_free (gd1_mgmt_commit_op_req *req)
{
        if (!req)
                return;

        GF_FREE (req->buf.buf_val);
        GF_FREE (req);
}

static void
gd_brick_op_req_free (gd1_mgmt_brick_op_req *req)
{
        if (!req)
                return;

        if (strcmp (req->name, "") != 0)
                GF_FREE (req->name);
        GF_FREE (req->input.input_val);
        GF_FREE (req);
}

int
gd_syncop_submit_request (struct rpc_clnt *rpc, void *req, void *local,
                          void *cookie, rpc_clnt_prog_t *prog, int procnum,
                          fop_cbk_fn_t cbkfn, xdrproc_t xdrproc)
{
        int            ret      = -1;
        struct iobuf  *iobuf    = NULL;
        struct iobref *iobref   = NULL;
        int            count    = 0;
        struct iovec   iov      = {0, };
        ssize_t        req_size = 0;
        call_frame_t  *frame    = NULL;

        GF_ASSERT (rpc);
        if (!req)
                goto out;

        req_size = xdr_sizeof (xdrproc, req);
        iobuf = iobuf_get2 (rpc->ctx->iobuf_pool, req_size);
        if (!iobuf)
                goto out;

        iobref = iobref_new ();
        if (!iobref)
                goto out;

        frame = create_frame (THIS, THIS->ctx->pool);
        if (!frame)
                goto out;

        iobref_add (iobref, iobuf);

        iov.iov_base = iobuf->ptr;
        iov.iov_len  = iobuf_pagesize (iobuf);

        /* Create the xdr payload */
        ret = xdr_serialize_generic (iov, req, xdrproc);
        if (ret == -1)
                goto out;

        iov.iov_len = ret;
        count = 1;

        frame->local = local;
        frame->cookie = cookie;

        /* Send the msg */
        ret = rpc_clnt_submit (rpc, prog, procnum, cbkfn,
                               &iov, count, NULL, 0, iobref,
                               frame, NULL, 0, NULL, 0, NULL);

        /* TODO: do we need to start ping also? */

out:
        iobref_unref (iobref);
        iobuf_unref (iobuf);

        return ret;
}

/* Defined in glusterd-rpc-ops.c */
extern struct rpc_clnt_program gd_mgmt_prog;
extern struct rpc_clnt_program gd_brick_prog;

static int
glusterd_syncop_aggr_rsp_dict (glusterd_op_t op, dict_t *aggr, dict_t *rsp)
{
        int ret = 0;

        switch (op) {
        case GD_OP_REPLACE_BRICK:
                ret = glusterd_rb_use_rsp_dict (aggr, rsp);
                if (ret)
                        goto out;
        break;

        case GD_OP_SYNC_VOLUME:
                ret = glusterd_sync_use_rsp_dict (aggr, rsp);
                if (ret)
                        goto out;
        break;

        case GD_OP_PROFILE_VOLUME:
                ret = glusterd_profile_volume_use_rsp_dict (aggr, rsp);
                if (ret)
                        goto out;
        break;

        case GD_OP_GSYNC_SET:
                ret = glusterd_gsync_use_rsp_dict (aggr, rsp, NULL);
                if (ret)
                        goto out;
        break;

        case GD_OP_STATUS_VOLUME:
                ret = glusterd_volume_status_copy_to_op_ctx_dict (aggr, rsp);
                if (ret)
                        goto out;
        break;

        case GD_OP_REBALANCE:
        case GD_OP_DEFRAG_BRICK_VOLUME:
                ret = glusterd_volume_rebalance_use_rsp_dict (aggr, rsp);
                if (ret)
                        goto out;
        break;

        case GD_OP_HEAL_VOLUME:
                ret = glusterd_volume_heal_use_rsp_dict (aggr, rsp);
                if (ret)
                        goto out;

        break;

        case GD_OP_QUOTA:
        case GD_OP_CLEARLOCKS_VOLUME:
                ret = glusterd_use_rsp_dict (aggr, rsp);
                if (ret)
                        goto out;

        break;

        default:
        break;
        }
out:
        return ret;
}

int32_t
_gd_syncop_mgmt_lock_cbk (struct rpc_req *req, struct iovec *iov,
                          int count, void *myframe)
{
        int                         ret         = -1;
        struct syncargs             *args       = NULL;
        glusterd_peerinfo_t         *peerinfo   = NULL;
        gd1_mgmt_cluster_lock_rsp   rsp         = {{0},};
        call_frame_t                *frame      = NULL;
        int                         op_ret      = -1;
        int                         op_errno    = -1;

        frame  = myframe;
        args   = frame->local;
        peerinfo = frame->cookie;
        frame->local = NULL;
        frame->cookie = NULL;

        if (-1 == req->rpc_status) {
                op_errno = ENOTCONN;
                goto out;
        }

        ret = xdr_to_generic (*iov, &rsp,
                              (xdrproc_t)xdr_gd1_mgmt_cluster_lock_rsp);
        if (ret < 0)
                goto out;

        uuid_copy (args->uuid, rsp.uuid);

        /* Set peer as locked, so we unlock only the locked peers */
        if (rsp.op_ret == 0)
                peerinfo->locked = _gf_true;
        op_ret = rsp.op_ret;
        op_errno = rsp.op_errno;
out:
        gd_collate_errors (args, op_ret, op_errno, NULL);
        STACK_DESTROY (frame->root);
        synctask_barrier_wake(args);
        return 0;
}

int32_t
gd_syncop_mgmt_lock_cbk (struct rpc_req *req, struct iovec *iov, int count,
                         void *myframe)
{
        return glusterd_big_locked_cbk (req, iov, count, myframe,
                                        _gd_syncop_mgmt_lock_cbk);
}

int
gd_syncop_mgmt_lock (glusterd_peerinfo_t *peerinfo, struct syncargs *args,
                     uuid_t my_uuid, uuid_t recv_uuid)
{
        int                       ret = -1;
        gd1_mgmt_cluster_lock_req req  = {{0},};
        glusterd_conf_t           *conf = THIS->private;

        uuid_copy (req.uuid, my_uuid);
        synclock_unlock (&conf->big_lock);
        ret = gd_syncop_submit_request (peerinfo->rpc, &req, args, peerinfo,
                                        &gd_mgmt_prog,
                                        GLUSTERD_MGMT_CLUSTER_LOCK,
                                        gd_syncop_mgmt_lock_cbk,
                                        (xdrproc_t) xdr_gd1_mgmt_cluster_lock_req);
        synclock_lock (&conf->big_lock);
        return ret;
}

int32_t
_gd_syncop_mgmt_unlock_cbk (struct rpc_req *req, struct iovec *iov,
                            int count, void *myframe)
{
        int                         ret         = -1;
        struct syncargs             *args       = NULL;
        glusterd_peerinfo_t         *peerinfo   = NULL;
        gd1_mgmt_cluster_unlock_rsp rsp         = {{0},};
        call_frame_t                *frame      = NULL;
        int                         op_ret      = -1;
        int                         op_errno    = -1;

        frame = myframe;
        args  = frame->local;
        peerinfo = frame->cookie;
        frame->local = NULL;

        if (-1 == req->rpc_status) {
                op_errno = ENOTCONN;
                goto out;
        }

        ret = xdr_to_generic (*iov, &rsp,
                              (xdrproc_t)xdr_gd1_mgmt_cluster_unlock_rsp);
        if (ret < 0)
                goto out;

        uuid_copy (args->uuid, rsp.uuid);

        peerinfo->locked = _gf_false;
        op_ret = rsp.op_ret;
        op_errno = rsp.op_errno;
out:
        gd_collate_errors (args, op_ret, op_errno, NULL);
        STACK_DESTROY (frame->root);
        synctask_barrier_wake(args);
        return 0;
}

int32_t
gd_syncop_mgmt_unlock_cbk (struct rpc_req *req, struct iovec *iov,
                           int count, void *myframe)
{
        return glusterd_big_locked_cbk (req, iov, count, myframe,
                                        _gd_syncop_mgmt_unlock_cbk);
}


int
gd_syncop_mgmt_unlock (glusterd_peerinfo_t *peerinfo, struct syncargs *args,
                       uuid_t my_uuid, uuid_t recv_uuid)
{
        int                         ret     = -1;
        gd1_mgmt_cluster_unlock_req req     = {{0},};
        glusterd_conf_t             *conf   = THIS->private;

        uuid_copy (req.uuid, my_uuid);
        synclock_unlock (&conf->big_lock);
        ret = gd_syncop_submit_request (peerinfo->rpc, &req, args, peerinfo,
                                        &gd_mgmt_prog,
                                        GLUSTERD_MGMT_CLUSTER_UNLOCK,
                                        gd_syncop_mgmt_unlock_cbk,
                                        (xdrproc_t) xdr_gd1_mgmt_cluster_lock_req);
        synclock_lock (&conf->big_lock);
        return ret;
}

int32_t
_gd_syncop_stage_op_cbk (struct rpc_req *req, struct iovec *iov,
                         int count, void *myframe)
{
        int                         ret         = -1;
        gd1_mgmt_stage_op_rsp       rsp         = {{0},};
        struct syncargs             *args       = NULL;
        xlator_t                    *this       = NULL;
        dict_t                      *rsp_dict   = NULL;
        call_frame_t                *frame      = NULL;
        int                         op_ret      = -1;
        int                         op_errno    = -1;

        this  = THIS;
        frame = myframe;
        args  = frame->local;
        frame->local = NULL;

        if (-1 == req->rpc_status) {
                op_errno = ENOTCONN;
                goto out;
        }

        ret = xdr_to_generic (*iov, &rsp,
                              (xdrproc_t)xdr_gd1_mgmt_stage_op_rsp);
        if (ret < 0)
                goto out;

        if (rsp.dict.dict_len) {
                /* Unserialize the dictionary */
                rsp_dict  = dict_new ();

                ret = dict_unserialize (rsp.dict.dict_val,
                                        rsp.dict.dict_len,
                                        &rsp_dict);
                if (ret < 0) {
                        GF_FREE (rsp.dict.dict_val);
                        goto out;
                } else {
                        rsp_dict->extra_stdfree = rsp.dict.dict_val;
                }
        }

        uuid_copy (args->uuid, rsp.uuid);
        if (rsp.op == GD_OP_REPLACE_BRICK) {
                pthread_mutex_lock (&args->lock_dict);
                {
                        ret = glusterd_syncop_aggr_rsp_dict (rsp.op, args->dict,
                                                             rsp_dict);
                        if (ret)
                                gf_log (this->name, GF_LOG_ERROR, "%s",
                                        "Failed to aggregate response from "
                                        " node/brick");
                }
                pthread_mutex_unlock (&args->lock_dict);
        }

        op_ret = rsp.op_ret;
        op_errno = rsp.op_errno;

out:
        gd_collate_errors (args, op_ret, op_errno, rsp.op_errstr);
        if (rsp_dict)
                dict_unref (rsp_dict);

        STACK_DESTROY (frame->root);
        synctask_barrier_wake(args);
        return 0;
}

int32_t
gd_syncop_stage_op_cbk (struct rpc_req *req, struct iovec *iov,
                        int count, void *myframe)
{
        return glusterd_big_locked_cbk (req, iov, count, myframe,
                                        _gd_syncop_stage_op_cbk);
}


int
gd_syncop_mgmt_stage_op (struct rpc_clnt *rpc, struct syncargs *args,
                         uuid_t my_uuid, uuid_t recv_uuid, int op,
                         dict_t *dict_out, dict_t *op_ctx)
{
        gd1_mgmt_stage_op_req *req  = NULL;
        glusterd_conf_t       *conf = THIS->private;
        int                   ret   = -1;

        req = GF_CALLOC (1, sizeof (*req), gf_gld_mt_mop_stage_req_t);
        if (!req)
                goto out;

        uuid_copy (req->uuid, my_uuid);
        req->op = op;

        ret = dict_allocate_and_serialize (dict_out,
                                           &req->buf.buf_val, &req->buf.buf_len);
        if (ret)
                goto out;

        synclock_unlock (&conf->big_lock);
        ret = gd_syncop_submit_request (rpc, req, args, NULL, &gd_mgmt_prog,
                                        GLUSTERD_MGMT_STAGE_OP,
                                        gd_syncop_stage_op_cbk,
                                        (xdrproc_t) xdr_gd1_mgmt_stage_op_req);
        synclock_lock (&conf->big_lock);
out:
        gd_stage_op_req_free (req);
        return ret;

}

int32_t
_gd_syncop_brick_op_cbk (struct rpc_req *req, struct iovec *iov,
                        int count, void *myframe)
{
        struct syncargs        *args  = NULL;
        gd1_mgmt_brick_op_rsp  rsp   = {0,};
        int                    ret   = -1;
        call_frame_t           *frame = NULL;

        frame = myframe;
        args = frame->local;
        frame->local = NULL;

        /* initialize */
        args->op_ret   = -1;
        args->op_errno = EINVAL;

        if (-1 == req->rpc_status) {
                args->op_errno = ENOTCONN;
                goto out;
        }

        ret = xdr_to_generic (*iov, &rsp,
                              (xdrproc_t)xdr_gd1_mgmt_brick_op_rsp);
        if (ret < 0)
                goto out;

        if (rsp.output.output_len) {
                args->dict  = dict_new ();
                if (!args->dict) {
                        ret = -1;
                        args->op_errno = ENOMEM;
                        goto out;
                }

                ret = dict_unserialize (rsp.output.output_val,
                                        rsp.output.output_len,
                                        &args->dict);
                if (ret < 0)
                        goto out;
        }

        args->op_ret = rsp.op_ret;
        args->op_errno = rsp.op_errno;
        args->errstr = gf_strdup (rsp.op_errstr);

out:
        if ((rsp.op_errstr) && (strcmp (rsp.op_errstr, "") != 0))
                free (rsp.op_errstr);
        free (rsp.output.output_val);

        STACK_DESTROY (frame->root);
        __wake (args);

        return 0;
}

int32_t
gd_syncop_brick_op_cbk (struct rpc_req *req, struct iovec *iov,
                        int count, void *myframe)
{
        return glusterd_big_locked_cbk (req, iov, count, myframe,
                                        _gd_syncop_brick_op_cbk);
}

int
gd_syncop_mgmt_brick_op (struct rpc_clnt *rpc, glusterd_pending_node_t *pnode,
                         int op, dict_t *dict_out, dict_t *op_ctx,
                         char **errstr)
{
        struct syncargs        args = {0, };
        gd1_mgmt_brick_op_req  *req  = NULL;
        int                    ret  = 0;
        xlator_t               *this = NULL;

        this = THIS;
        args.op_ret = -1;
        args.op_errno = ENOTCONN;

        if ((pnode->type == GD_NODE_NFS) ||
            ((pnode->type == GD_NODE_SHD) &&
            (op == GD_OP_STATUS_VOLUME))) {
                ret = glusterd_node_op_build_payload
                        (op, &req, dict_out);

        } else {
                ret = glusterd_brick_op_build_payload
                        (op, pnode->node, &req, dict_out);

        }

        if (ret)
                goto out;

        GD_SYNCOP (rpc, (&args), NULL, gd_syncop_brick_op_cbk, req,
                   &gd_brick_prog, req->op, xdr_gd1_mgmt_brick_op_req);

        if (args.errstr && errstr)
                *errstr = args.errstr;
        else
                GF_FREE (args.errstr);

        if (GD_OP_STATUS_VOLUME == op) {
                ret = dict_set_int32 (args.dict, "index", pnode->index);
                if (ret) {
                        gf_log (this->name, GF_LOG_ERROR,
                                "Error setting index on brick status"
                                " rsp dict");
                        args.op_ret = -1;
                        goto out;
                }
        }
        if (args.op_ret == 0)
                glusterd_handle_node_rsp (dict_out, pnode->node, op,
                                          args.dict, op_ctx, errstr,
                                          pnode->type);

out:
        errno = args.op_errno;
        if (args.dict)
                dict_unref (args.dict);
        gd_brick_op_req_free (req);
        return args.op_ret;

}

int32_t
_gd_syncop_commit_op_cbk (struct rpc_req *req, struct iovec *iov,
                          int count, void *myframe)
{
        int                         ret         = -1;
        gd1_mgmt_commit_op_rsp      rsp         = {{0},};
        struct syncargs             *args       = NULL;
        xlator_t                    *this       = NULL;
        dict_t                      *rsp_dict   = NULL;
        call_frame_t                *frame      = NULL;
        int                         op_ret      = -1;
        int                         op_errno    = -1;

        this  = THIS;
        frame = myframe;
        args  = frame->local;
        frame->local = NULL;

        if (-1 == req->rpc_status) {
                op_errno = ENOTCONN;
                goto out;
        }

        ret = xdr_to_generic (*iov, &rsp,
                              (xdrproc_t)xdr_gd1_mgmt_commit_op_rsp);
        if (ret < 0) {
                goto out;
        }

        if (rsp.dict.dict_len) {
                /* Unserialize the dictionary */
                rsp_dict  = dict_new ();

                ret = dict_unserialize (rsp.dict.dict_val,
                                        rsp.dict.dict_len,
                                        &rsp_dict);
                if (ret < 0) {
                        GF_FREE (rsp.dict.dict_val);
                        goto out;
                } else {
                        rsp_dict->extra_stdfree = rsp.dict.dict_val;
                }
        }

        uuid_copy (args->uuid, rsp.uuid);
        pthread_mutex_lock (&args->lock_dict);
        {
                ret = glusterd_syncop_aggr_rsp_dict (rsp.op, args->dict,
                                                     rsp_dict);
                if (ret)
                        gf_log (this->name, GF_LOG_ERROR, "%s",
                                "Failed to aggregate response from "
                                " node/brick");
        }
        pthread_mutex_unlock (&args->lock_dict);

        op_ret = rsp.op_ret;
        op_errno = rsp.op_errno;
out:
        gd_collate_errors (args, op_ret, op_errno, rsp.op_errstr);
        if (rsp_dict)
                dict_unref (rsp_dict);

        STACK_DESTROY (frame->root);
        synctask_barrier_wake(args);

        return 0;
}

int32_t
gd_syncop_commit_op_cbk (struct rpc_req *req, struct iovec *iov,
                         int count, void *myframe)
{
        return glusterd_big_locked_cbk (req, iov, count, myframe,
                                        _gd_syncop_commit_op_cbk);
}


int
gd_syncop_mgmt_commit_op (struct rpc_clnt *rpc, struct syncargs *args,
                          uuid_t my_uuid, uuid_t recv_uuid,
                          int op, dict_t *dict_out, dict_t *op_ctx)
{
        glusterd_conf_t        *conf = THIS->private;
        gd1_mgmt_commit_op_req *req  = NULL;
        int                    ret  = -1;

        req = GF_CALLOC (1, sizeof (*req), gf_gld_mt_mop_commit_req_t);
        if (!req)
                goto out;

        uuid_copy (req->uuid, my_uuid);
        req->op = op;

        ret = dict_allocate_and_serialize (dict_out,
                                           &req->buf.buf_val, &req->buf.buf_len);
        if (ret)
                goto out;

        synclock_unlock (&conf->big_lock);
        ret = gd_syncop_submit_request (rpc, req, args, NULL, &gd_mgmt_prog,
                                        GLUSTERD_MGMT_COMMIT_OP ,
                                        gd_syncop_commit_op_cbk,
                                        (xdrproc_t) xdr_gd1_mgmt_commit_op_req);
        synclock_lock (&conf->big_lock);
out:
        gd_commit_op_req_free (req);
        return ret;
}


int
gd_build_peers_list (struct list_head *peers, struct list_head *xact_peers,
                     glusterd_op_t op)
{
        glusterd_peerinfo_t *peerinfo = NULL;
        int                 npeers      = 0;

        list_for_each_entry (peerinfo, peers, uuid_list) {
                if (!peerinfo->connected)
                        continue;
                if (op != GD_OP_SYNC_VOLUME &&
                    peerinfo->state.state != GD_FRIEND_STATE_BEFRIENDED)
                        continue;

                list_add_tail (&peerinfo->op_peers_list, xact_peers);
                npeers++;
        }
        return npeers;
}

int
gd_lock_op_phase (struct list_head *peers, glusterd_op_t op, dict_t *op_ctx,
                  char **op_errstr, int npeers)
{
        int                 ret         = -1;
        int                 peer_cnt    = 0;
        uuid_t              peer_uuid   = {0};
        xlator_t            *this       = NULL;
        glusterd_peerinfo_t *peerinfo   = NULL;
        struct syncargs     args        = {0};

        if (!npeers) {
                ret = 0;
                goto out;
        }

        this = THIS;
        synctask_barrier_init((&args));
        peer_cnt = 0;
        list_for_each_entry (peerinfo, peers, op_peers_list) {
                /* Reset lock status */
                peerinfo->locked = _gf_false;
                gd_syncop_mgmt_lock (peerinfo, &args, MY_UUID, peer_uuid);
                peer_cnt++;
        }
        gd_synctask_barrier_wait((&args), peer_cnt);
        ret = args.op_ret;
        if (ret) {
                gf_asprintf (op_errstr, "Another transaction could be "
                             "in progress. Please try again after "
                             "sometime.");
                gf_log (this->name, GF_LOG_ERROR, "Failed to acquire lock");
                goto out;
        }

        ret = 0;
out:
        return ret;
}

int
gd_stage_op_phase (struct list_head *peers, glusterd_op_t op, dict_t *op_ctx,
                   dict_t *req_dict, char **op_errstr, int npeers)
{
        int                 ret             = -1;
        int                 peer_cnt           = 0;
        dict_t              *rsp_dict       = NULL;
        char                *hostname       = NULL;
        xlator_t            *this           = NULL;
        glusterd_peerinfo_t *peerinfo       = NULL;
        uuid_t              tmp_uuid        = {0};
        char                *errstr         = NULL;
        struct syncargs     args            = {0};

        this = THIS;
        rsp_dict = dict_new ();
        if (!rsp_dict)
                goto out;

        ret = glusterd_op_stage_validate (op, req_dict, op_errstr, rsp_dict);
        if (ret) {
                hostname = "localhost";
                goto stage_done;
        }

        if ((op == GD_OP_REPLACE_BRICK)) {
                ret = glusterd_syncop_aggr_rsp_dict (op, op_ctx, rsp_dict);
                if (ret) {
                        gf_log (this->name, GF_LOG_ERROR, "%s",
                                "Failed to aggregate response from node/brick");
                        goto out;
                }
        }
        dict_unref (rsp_dict);
        rsp_dict = NULL;

stage_done:
        if (ret) {
                gf_log (this->name, GF_LOG_ERROR, LOGSTR_STAGE_FAIL,
                        gd_op_list[op], hostname, (*op_errstr) ? ":" : " ",
                        (*op_errstr) ? *op_errstr : " ");
                if (*op_errstr == NULL)
                        gf_asprintf (op_errstr, OPERRSTR_STAGE_FAIL, hostname);
                goto out;
        }

        if (!npeers) {
                ret = 0;
                goto out;
        }

        gd_syncargs_init (&args, op_ctx);
        synctask_barrier_init((&args));
        peer_cnt = 0;
        list_for_each_entry (peerinfo, peers, op_peers_list) {
                ret = gd_syncop_mgmt_stage_op (peerinfo->rpc, &args,
                                               MY_UUID, tmp_uuid,
                                               op, req_dict, op_ctx);
                peer_cnt++;
        }
        gd_synctask_barrier_wait((&args), peer_cnt);
        ret = args.op_ret;
        if (dict_get_str (op_ctx, "errstr", &errstr) == 0)
                *op_errstr = gf_strdup (errstr);
        else if (args.errstr)
                *op_errstr = gf_strdup (args.errstr);

out:
        if (rsp_dict)
                dict_unref (rsp_dict);
        return ret;
}

int
gd_commit_op_phase (struct list_head *peers, glusterd_op_t op, dict_t *op_ctx,
                    dict_t *req_dict, char **op_errstr, int npeers)
{
        dict_t              *rsp_dict       = NULL;
        int                 peer_cnt           = -1;
        int                 ret             = -1;
        char                *hostname       = NULL;
        glusterd_peerinfo_t *peerinfo       = NULL;
        xlator_t            *this           = NULL;
        uuid_t              tmp_uuid        = {0};
        char                *errstr         = NULL;
        struct syncargs     args            = {0};

        this = THIS;
        rsp_dict = dict_new ();
        if (!rsp_dict) {
                ret = -1;
                goto out;
        }

        ret = glusterd_op_commit_perform (op, req_dict, op_errstr, rsp_dict);
        if (ret) {
                hostname = "localhost";
                goto commit_done;
        }
        if (op != GD_OP_SYNC_VOLUME) {
                ret =  glusterd_syncop_aggr_rsp_dict (op, op_ctx, rsp_dict);
                if (ret) {
                        gf_log (this->name, GF_LOG_ERROR, "%s",
                                "Failed to aggregate response "
                                "from node/brick");
                        goto out;
                }
        }
        dict_unref (rsp_dict);
        rsp_dict = NULL;

commit_done:
        if (ret) {
                gf_log (this->name, GF_LOG_ERROR, LOGSTR_COMMIT_FAIL,
                        gd_op_list[op], hostname, (*op_errstr) ? ":" : " ",
                        (*op_errstr) ? *op_errstr : " ");
                if (*op_errstr == NULL)
                        gf_asprintf (op_errstr, OPERRSTR_COMMIT_FAIL,
                                     hostname);
                goto out;
        }

        if (!npeers) {
                ret = 0;
                goto out;
        }
        gd_syncargs_init (&args, op_ctx);
        synctask_barrier_init((&args));
        peer_cnt = 0;
        list_for_each_entry (peerinfo, peers, op_peers_list) {
                ret = gd_syncop_mgmt_commit_op (peerinfo->rpc, &args,
                                                MY_UUID, tmp_uuid,
                                                op, req_dict, op_ctx);
                peer_cnt++;
        }
        gd_synctask_barrier_wait((&args), peer_cnt);
        ret = args.op_ret;
        if (dict_get_str (op_ctx, "errstr", &errstr) == 0)
                *op_errstr = gf_strdup (errstr);
        else if (args.errstr)
                *op_errstr = gf_strdup (args.errstr);

out:
        if (!ret)
                glusterd_op_modify_op_ctx (op, op_ctx);

        if (rsp_dict)
                dict_unref (rsp_dict);
        return ret;
}

int
gd_unlock_op_phase (struct list_head *peers, glusterd_op_t op, int op_ret,
                    rpcsvc_request_t *req, dict_t *op_ctx, char *op_errstr,
                    int npeers)
{
        glusterd_peerinfo_t *peerinfo   = NULL;
        glusterd_peerinfo_t *tmp        = NULL;
        uuid_t              tmp_uuid    = {0};
        int                 peer_cnt       = 0;
        int                 ret         = -1;
        xlator_t            *this       = NULL;
        struct syncargs     args        = {0};

        if (!npeers) {
                ret = 0;
                goto out;
        }

        this = THIS;
        synctask_barrier_init((&args));
        peer_cnt = 0;
        list_for_each_entry_safe (peerinfo, tmp, peers, op_peers_list) {
                /* Only unlock peers that were locked */
                if (peerinfo->locked) {
                        gd_syncop_mgmt_unlock (peerinfo, &args, MY_UUID,
                                               tmp_uuid);
                        peer_cnt++;
                }

                list_del_init (&peerinfo->op_peers_list);
        }
        gd_synctask_barrier_wait((&args), peer_cnt);
        ret = args.op_ret;
        if (ret) {
                gf_log (this->name, GF_LOG_ERROR, "Failed to unlock "
                        "on some peer(s)");
        }

out:
        glusterd_op_send_cli_response (op, op_ret, 0, req, op_ctx, op_errstr);
        glusterd_op_clear_op (op);
        glusterd_unlock (MY_UUID);

        return 0;
}

int
gd_get_brick_count (struct list_head *bricks)
{
        glusterd_pending_node_t *pending_node = NULL;
        int                     npeers        = 0;
        list_for_each_entry (pending_node, bricks, list) {
                npeers++;
        }
        return npeers;
}

int
gd_brick_op_phase (glusterd_op_t op, dict_t *op_ctx, dict_t *req_dict, char **op_errstr)
{
        glusterd_pending_node_t *pending_node = NULL;
        struct list_head        selected = {0,};
        xlator_t                *this = NULL;
        int                     brick_count = 0;
        int                     ret = -1;
        rpc_clnt_t              *rpc = NULL;
        dict_t                  *rsp_dict = NULL;
        glusterd_conf_t         *conf = NULL;

        this = THIS;
        conf = this->private;
        rsp_dict = dict_new ();
        if (!rsp_dict) {
                ret = -1;
                goto out;
        }

        INIT_LIST_HEAD (&selected);
        ret = glusterd_op_bricks_select (op, req_dict, op_errstr, &selected, rsp_dict);
        if (ret) {
                gf_log (this->name, GF_LOG_ERROR, "%s",
                       (*op_errstr)? *op_errstr: "Brick op failed. Check "
                       "glusterd log file for more details.");
                goto out;
        }

        if (op == GD_OP_HEAL_VOLUME) {
                ret = glusterd_syncop_aggr_rsp_dict (op, op_ctx, rsp_dict);
                if (ret)
                        goto out;
        }
        dict_unref (rsp_dict);
        rsp_dict = NULL;

        brick_count = 0;
        list_for_each_entry (pending_node, &selected, list) {
                rpc = glusterd_pending_node_get_rpc (pending_node);
                if (!rpc) {
                        if (pending_node->type == GD_NODE_REBALANCE) {
                                ret = 0;
                                glusterd_defrag_volume_node_rsp (req_dict,
                                                                 NULL, op_ctx);
                                goto out;
                        }

                        ret = -1;
                        gf_log (this->name, GF_LOG_ERROR, "Brick Op failed "
                                "due to rpc failure.");
                        goto out;
                }
                ret = gd_syncop_mgmt_brick_op (rpc, pending_node, op, req_dict,
                                               op_ctx, op_errstr);
                if (ret)
                        goto out;

                brick_count++;
        }

        ret = 0;
out:
        if (rsp_dict)
                dict_unref (rsp_dict);
        gf_log (this->name, GF_LOG_DEBUG, "Sent op req to %d bricks",
                brick_count);
        return ret;
}

void
gd_sync_task_begin (dict_t *op_ctx, rpcsvc_request_t * req)
{
        int                         ret             = -1;
        int                         npeers          = 0;
        dict_t                      *req_dict       = NULL;
        glusterd_conf_t             *conf           = NULL;
        glusterd_op_t               op              = 0;
        int32_t                     tmp_op          = 0;
        char                        *op_errstr      = NULL;
        xlator_t                    *this           = NULL;

        this = THIS;
        GF_ASSERT (this);
        conf = this->private;
        GF_ASSERT (conf);

        ret = dict_get_int32 (op_ctx, GD_SYNC_OPCODE_KEY, &tmp_op);
        if (ret) {
                gf_log (this->name, GF_LOG_ERROR, "Failed to get volume "
                        "operation");
                goto out;
        }

        op = tmp_op;
        ret = glusterd_lock (MY_UUID);
        if (ret) {
                gf_log (this->name, GF_LOG_ERROR, "Unable to acquire lock");
                gf_asprintf (&op_errstr, "Another transaction is in progress. "
                             "Please try again after sometime.");
                goto out;
        }

        /* storing op globally to access in synctask code paths
         * This is still acceptable, as we are performing this under
         * the 'cluster' lock*/
        glusterd_op_set_op  (op);
        INIT_LIST_HEAD (&conf->xaction_peers);
        npeers = gd_build_peers_list  (&conf->peers, &conf->xaction_peers, op);

        ret = gd_lock_op_phase (&conf->xaction_peers, op, op_ctx, &op_errstr, npeers);
        if (ret)
                goto out;

        ret = glusterd_op_build_payload (&req_dict, &op_errstr, op_ctx);
        if (ret) {
                gf_log (this->name, GF_LOG_ERROR, LOGSTR_BUILD_PAYLOAD,
                        gd_op_list[op]);
                if (op_errstr == NULL)
                        gf_asprintf (&op_errstr, OPERRSTR_BUILD_PAYLOAD);
                goto out;
        }

        ret = gd_stage_op_phase (&conf->xaction_peers, op, op_ctx, req_dict,
                                 &op_errstr, npeers);
        if (ret)
                goto out;

        ret = gd_brick_op_phase (op, op_ctx, req_dict, &op_errstr);
        if (ret)
                goto out;

        ret = gd_commit_op_phase (&conf->xaction_peers, op, op_ctx, req_dict,
                                  &op_errstr, npeers);
        if (ret)
                goto out;

        ret = 0;
out:
        (void) gd_unlock_op_phase (&conf->xaction_peers, op, ret, req,
                                   op_ctx, op_errstr, npeers);

        if (req_dict)
                dict_unref (req_dict);

        if (op_errstr)
                GF_FREE (op_errstr);

        return;
}

int32_t
glusterd_op_begin_synctask (rpcsvc_request_t *req, glusterd_op_t op,
                            void *dict)
{
        int              ret = 0;

        ret = dict_set_int32 (dict, GD_SYNC_OPCODE_KEY, op);
        if (ret) {
                gf_log (THIS->name, GF_LOG_ERROR,
                        "dict set failed for setting operations");
                goto out;
        }

        gd_sync_task_begin (dict, req);
        ret = 0;
out:

        return ret;
}