1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471 | // Copyright (C) 2018-2024 Internet Systems Consortium, Inc. ("ISC")
//
// This Source Code Form is subject to the terms of the Mozilla Public
// License, v. 2.0. If a copy of the MPL was not distributed with this
// file, You can obtain one at http://mozilla.org/MPL/2.0/.
#include <config.h>
#include <command_creator.h><--- Include file: not found. Please note: Cppcheck does not need standard library headers to get proper results.
#include <ha_log.h><--- Include file: not found. Please note: Cppcheck does not need standard library headers to get proper results.
#include <ha_service.h><--- Include file: not found. Please note: Cppcheck does not need standard library headers to get proper results.
#include <ha_service_states.h><--- Include file: not found. Please note: Cppcheck does not need standard library headers to get proper results.
#include <cc/command_interpreter.h>
#include <cc/data.h>
#include <config/cmd_response_creator.h>
#include <config/timeouts.h>
#include <dhcp/iface_mgr.h>
#include <dhcpsrv/cfgmgr.h>
#include <dhcpsrv/lease_mgr.h>
#include <dhcpsrv/lease_mgr_factory.h>
#include <exceptions/exceptions.h>
#include <http/date_time.h>
#include <http/response_json.h>
#include <http/post_request_json.h>
#include <util/boost_time_utils.h>
#include <util/multi_threading_mgr.h>
#include <util/stopwatch.h>
#include <boost/pointer_cast.hpp><--- Include file: not found. Please note: Cppcheck does not need standard library headers to get proper results.
#include <boost/make_shared.hpp><--- Include file: not found. Please note: Cppcheck does not need standard library headers to get proper results.
#include <boost/weak_ptr.hpp><--- Include file: not found. Please note: Cppcheck does not need standard library headers to get proper results.
#include <functional><--- Include file: not found. Please note: Cppcheck does not need standard library headers to get proper results.
#include <sstream><--- Include file: not found. Please note: Cppcheck does not need standard library headers to get proper results.
using namespace isc::asiolink;
using namespace isc::config;
using namespace isc::data;
using namespace isc::dhcp;
using namespace isc::hooks;
using namespace isc::http;
using namespace isc::log;
using namespace isc::util;
namespace ph = std::placeholders;
namespace {
/// @brief Exception thrown when command sent to the partner is unsupported.
class CommandUnsupportedError : public CtrlChannelError {
public:
CommandUnsupportedError(const char* file, size_t line, const char* what) :
CtrlChannelError(file, line, what) {}
};
/// @brief Exception thrown when conflict status code has been returned.
class ConflictError : public CtrlChannelError {
public:
ConflictError(const char* file, size_t line, const char* what) :
CtrlChannelError(file, line, what) {}
};
}
namespace isc {
namespace ha {
const int HAService::HA_HEARTBEAT_COMPLETE_EVT;
const int HAService::HA_LEASE_UPDATES_COMPLETE_EVT;
const int HAService::HA_SYNCING_FAILED_EVT;
const int HAService::HA_SYNCING_SUCCEEDED_EVT;
const int HAService::HA_MAINTENANCE_NOTIFY_EVT;
const int HAService::HA_MAINTENANCE_START_EVT;
const int HAService::HA_MAINTENANCE_CANCEL_EVT;
const int HAService::HA_CONTROL_RESULT_MAINTENANCE_NOT_ALLOWED;
const int HAService::HA_SYNCED_PARTNER_UNAVAILABLE_EVT;
const int HAService::HA_WAITING_TO_TERMINATED_ST_DELAY_MINUTES;
HAService::HAService(const unsigned int id, const IOServicePtr& io_service,
const NetworkStatePtr& network_state, const HAConfigPtr& config,
const HAServerType& server_type)
: id_(id), io_service_(io_service), network_state_(network_state), config_(config),
server_type_(server_type), client_(), listener_(), communication_state_(),
query_filter_(config), lease_sync_filter_(server_type, config), mutex_(),
pending_requests_(), lease_update_backlog_(config->getDelayedUpdatesLimit()),
sync_complete_notified_(false) {
if (server_type == HAServerType::DHCPv4) {
communication_state_.reset(new CommunicationState4(io_service_, config));
} else {
communication_state_.reset(new CommunicationState6(io_service_, config));
}
network_state_->enableService(getLocalOrigin());
startModel(HA_WAITING_ST);
// Create the client and(or) listener as appropriate.
if (!config_->getEnableMultiThreading()) {
// Not configured for multi-threading, start a client in ST mode.
client_.reset(new HttpClient(io_service_, false));
} else {
// Create an MT-mode client.
client_.reset(new HttpClient(io_service_, true,
config_->getHttpClientThreads(), true));
// If we're configured to use our own listener create and start it.
if (config_->getHttpDedicatedListener()) {
// Get the server address and port from this server's URL.
auto my_url = config_->getThisServerConfig()->getUrl();
IOAddress server_address(IOAddress::IPV4_ZERO_ADDRESS());
try {
// Since we do not currently support hostname resolution,
// we need to make sure we have an IP address here.
server_address = IOAddress(my_url.getStrippedHostname());
} catch (const std::exception& ex) {
isc_throw(Unexpected, "server Url:" << my_url.getStrippedHostname()
<< " is not a valid IP address");
}
// Fetch how many threads the listener will use.
uint32_t listener_threads = config_->getHttpListenerThreads();
// Fetch the TLS context.
auto tls_context = config_->getThisServerConfig()->getTlsContext();
// Instantiate the listener.
listener_.reset(new CmdHttpListener(server_address, my_url.getPort(),
listener_threads, tls_context));
// Set the command filter when enabled.
if (config_->getRestrictCommands()) {
if (server_type == HAServerType::DHCPv4) {
CmdResponseCreator::command_accept_list_ =
CommandCreator::ha_commands4_;
} else {
CmdResponseCreator::command_accept_list_ =
CommandCreator::ha_commands6_;
}
}
}
}
LOG_INFO(ha_logger, HA_SERVICE_STARTED)
.arg(config_->getThisServerName())
.arg(HAConfig::HAModeToString(config->getHAMode()))
.arg(HAConfig::PeerConfig::roleToString(config->getThisServerConfig()->getRole()));
}
HAService::~HAService() {
// Stop client and/or listener.
stopClientAndListener();
network_state_->enableService(getLocalOrigin());
}
std::string
HAService::getCSCallbacksSetName() const {
std::ostringstream s;
s << "HA_MT_" << id_;
return (s.str());
}
void
HAService::defineEvents() {
StateModel::defineEvents();
defineEvent(HA_HEARTBEAT_COMPLETE_EVT, "HA_HEARTBEAT_COMPLETE_EVT");
defineEvent(HA_LEASE_UPDATES_COMPLETE_EVT, "HA_LEASE_UPDATES_COMPLETE_EVT");
defineEvent(HA_SYNCING_FAILED_EVT, "HA_SYNCING_FAILED_EVT");
defineEvent(HA_SYNCING_SUCCEEDED_EVT, "HA_SYNCING_SUCCEEDED_EVT");
defineEvent(HA_MAINTENANCE_NOTIFY_EVT, "HA_MAINTENANCE_NOTIFY_EVT");
defineEvent(HA_MAINTENANCE_START_EVT, "HA_MAINTENANCE_START_EVT");
defineEvent(HA_MAINTENANCE_CANCEL_EVT, "HA_MAINTENANCE_CANCEL_EVT");
defineEvent(HA_SYNCED_PARTNER_UNAVAILABLE_EVT, "HA_SYNCED_PARTNER_UNAVAILABLE_EVT");
}
void
HAService::verifyEvents() {
StateModel::verifyEvents();
getEvent(HA_HEARTBEAT_COMPLETE_EVT);
getEvent(HA_LEASE_UPDATES_COMPLETE_EVT);
getEvent(HA_SYNCING_FAILED_EVT);
getEvent(HA_SYNCING_SUCCEEDED_EVT);
getEvent(HA_MAINTENANCE_NOTIFY_EVT);
getEvent(HA_MAINTENANCE_START_EVT);
getEvent(HA_MAINTENANCE_CANCEL_EVT);
getEvent(HA_SYNCED_PARTNER_UNAVAILABLE_EVT);
}
void
HAService::defineStates() {
StateModel::defineStates();
defineState(HA_BACKUP_ST, stateToString(HA_BACKUP_ST),
std::bind(&HAService::backupStateHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_BACKUP_ST)->getPausing());
defineState(HA_COMMUNICATION_RECOVERY_ST, stateToString(HA_COMMUNICATION_RECOVERY_ST),
std::bind(&HAService::communicationRecoveryHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_COMMUNICATION_RECOVERY_ST)->getPausing());
defineState(HA_HOT_STANDBY_ST, stateToString(HA_HOT_STANDBY_ST),
std::bind(&HAService::normalStateHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_HOT_STANDBY_ST)->getPausing());
defineState(HA_LOAD_BALANCING_ST, stateToString(HA_LOAD_BALANCING_ST),
std::bind(&HAService::normalStateHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_LOAD_BALANCING_ST)->getPausing());
defineState(HA_IN_MAINTENANCE_ST, stateToString(HA_IN_MAINTENANCE_ST),
std::bind(&HAService::inMaintenanceStateHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_IN_MAINTENANCE_ST)->getPausing());
defineState(HA_PARTNER_DOWN_ST, stateToString(HA_PARTNER_DOWN_ST),
std::bind(&HAService::partnerDownStateHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_PARTNER_DOWN_ST)->getPausing());
defineState(HA_PARTNER_IN_MAINTENANCE_ST, stateToString(HA_PARTNER_IN_MAINTENANCE_ST),
std::bind(&HAService::partnerInMaintenanceStateHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_PARTNER_IN_MAINTENANCE_ST)->getPausing());
defineState(HA_PASSIVE_BACKUP_ST, stateToString(HA_PASSIVE_BACKUP_ST),
std::bind(&HAService::passiveBackupStateHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_PASSIVE_BACKUP_ST)->getPausing());
defineState(HA_READY_ST, stateToString(HA_READY_ST),
std::bind(&HAService::readyStateHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_READY_ST)->getPausing());
defineState(HA_SYNCING_ST, stateToString(HA_SYNCING_ST),
std::bind(&HAService::syncingStateHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_SYNCING_ST)->getPausing());
defineState(HA_TERMINATED_ST, stateToString(HA_TERMINATED_ST),
std::bind(&HAService::terminatedStateHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_TERMINATED_ST)->getPausing());
defineState(HA_WAITING_ST, stateToString(HA_WAITING_ST),
std::bind(&HAService::waitingStateHandler, this),
config_->getStateMachineConfig()->getStateConfig(HA_WAITING_ST)->getPausing());
}
void
HAService::backupStateHandler() {
if (doOnEntry()) {
query_filter_.serveNoScopes();
adjustNetworkState();
// Log if the state machine is paused.
conditionalLogPausedState();
}
// There is nothing to do in that state. This server simply receives
// lease updates from the partners.
postNextEvent(NOP_EVT);
}
void
HAService::communicationRecoveryHandler() {
if (doOnEntry()) {
query_filter_.serveDefaultScopes();
adjustNetworkState();
// Log if the state machine is paused.
conditionalLogPausedState();
}
scheduleHeartbeat();
if (isMaintenanceCanceled() || isModelPaused()) {
postNextEvent(NOP_EVT);
// Check if the clock skew is still acceptable. If not, transition to
// the terminated state.
} else if (shouldTerminate()) {
verboseTransition(HA_TERMINATED_ST);
} else if (isPartnerStateInvalid()) {
verboseTransition(HA_WAITING_ST);
} else {
// Transitions based on the partner's state.
switch (communication_state_->getPartnerState()) {
case HA_IN_MAINTENANCE_ST:
verboseTransition(HA_PARTNER_IN_MAINTENANCE_ST);
break;
case HA_PARTNER_DOWN_ST:
verboseTransition(HA_WAITING_ST);
break;
case HA_PARTNER_IN_MAINTENANCE_ST:
verboseTransition(HA_IN_MAINTENANCE_ST);
break;
case HA_TERMINATED_ST:
verboseTransition(HA_TERMINATED_ST);
break;
case HA_UNAVAILABLE_ST:
if (shouldPartnerDown()) {
verboseTransition(HA_PARTNER_DOWN_ST);
} else {
postNextEvent(NOP_EVT);
}
break;
case HA_WAITING_ST:
case HA_SYNCING_ST:
case HA_READY_ST:
// The partner seems to be waking up, perhaps after communication-recovery.
// If our backlog queue is overflown we need to synchronize our lease database.
// There is no need to send ha-reset to the partner because the partner is
// already synchronizing its lease database.
if (!communication_state_->isCommunicationInterrupted() &&
lease_update_backlog_.wasOverflown()) {
verboseTransition(HA_WAITING_ST);
} else {
// Backlog was not overflown, so there is no need to synchronize our
// lease database. Let's wait until our partner completes synchronization
// and transitions to the load-balancing state.
postNextEvent(NOP_EVT);
}
break;
default:
// If the communication is still interrupted, let's continue sitting
// in this state until it is resumed or until the transition to the
// partner-down state, depending on what happens first.
if (communication_state_->isCommunicationInterrupted()) {
postNextEvent(NOP_EVT);
break;
}
// The communication has been resumed. The partner server must be in a state
// in which it can receive outstanding lease updates we collected. The number of
// outstanding lease updates must not exceed the configured limit. Finally, the
// lease updates must be successfully sent. If that all works, we will transition
// to the normal operation.
if ((communication_state_->getPartnerState() == getNormalState()) ||
(communication_state_->getPartnerState() == HA_COMMUNICATION_RECOVERY_ST)) {
if (lease_update_backlog_.wasOverflown() || !sendLeaseUpdatesFromBacklog()) {
// If our lease backlog was overflown or we were unable to send lease
// updates to the partner we should notify the partner that it should
// synchronize the lease database. We do it by sending ha-reset command.
if (sendHAReset()) {
verboseTransition(HA_WAITING_ST);
}
break;
}
// The backlog was not overflown and we successfully sent our lease updates.
// We can now transition to the normal operation state. If the partner
// fails to send his outstanding lease updates to us it should send the
// ha-reset command to us.
verboseTransition(getNormalState());
break;
}
// The partner appears to be in unexpected state, we have exceeded the number
// of lease updates in a backlog or an attempt to send lease updates failed.
// In all these cases we follow plan B and transition to the waiting state.
// The server will then attempt to synchronize the entire lease database.
verboseTransition(HA_WAITING_ST);
}
}
// When exiting this state we must ensure that lease updates backlog is cleared.
if (doOnExit()) {
lease_update_backlog_.clear();
}
}
void
HAService::normalStateHandler() {
// If we are transitioning from another state, we have to define new
// serving scopes appropriate for the new state. We don't do it if
// we remain in this state.
if (doOnEntry()) {
query_filter_.serveDefaultScopes();
adjustNetworkState();
// Log if the state machine is paused.
conditionalLogPausedState();
}
scheduleHeartbeat();
if (isMaintenanceCanceled() || isModelPaused()) {
postNextEvent(NOP_EVT);
return;
}
// Check if the clock skew is still acceptable. If not, transition to
// the terminated state.
if (shouldTerminate()) {
verboseTransition(HA_TERMINATED_ST);
return;
}
// Check if the partner state is valid per current configuration. If it is
// in an invalid state let's transition to the waiting state and stay there
// until the configuration is corrected.
if (isPartnerStateInvalid()) {
verboseTransition(HA_WAITING_ST);
return;
}
switch (communication_state_->getPartnerState()) {
case HA_IN_MAINTENANCE_ST:
verboseTransition(HA_PARTNER_IN_MAINTENANCE_ST);
break;
case HA_PARTNER_DOWN_ST:
verboseTransition(HA_WAITING_ST);
break;
case HA_PARTNER_IN_MAINTENANCE_ST:
verboseTransition(HA_IN_MAINTENANCE_ST);
break;
case HA_TERMINATED_ST:
verboseTransition(HA_TERMINATED_ST);
break;
case HA_UNAVAILABLE_ST:
if (shouldPartnerDown()) {
verboseTransition(HA_PARTNER_DOWN_ST);
} else if (config_->amAllowingCommRecovery()) {
verboseTransition(HA_COMMUNICATION_RECOVERY_ST);
} else {
postNextEvent(NOP_EVT);
}
break;
default:
postNextEvent(NOP_EVT);
}
if (doOnExit()) {
// Do nothing here but doOnExit() call clears the "on exit" flag
// when transitioning to the communication-recovery state. In that
// state we need this flag to be cleared.
}
}
void
HAService::inMaintenanceStateHandler() {
// If we are transitioning from another state, we have to define new
// serving scopes appropriate for the new state. We don't do it if
// we remain in this state.
if (doOnEntry()) {
// In this state the server remains silent and waits for being
// shutdown.
query_filter_.serveNoScopes();
adjustNetworkState();
// Log if the state machine is paused.
conditionalLogPausedState();
LOG_INFO(ha_logger, HA_MAINTENANCE_SHUTDOWN_SAFE)
.arg(config_->getThisServerName());
}
scheduleHeartbeat();
// We don't transition out of this state unless explicitly mandated
// by the administrator via a dedicated command which cancels
// the maintenance.
postNextEvent(NOP_EVT);
}
void
HAService::partnerDownStateHandler() {
// If we are transitioning from another state, we have to define new
// serving scopes appropriate for the new state. We don't do it if
// we remain in this state.
if (doOnEntry()) {
bool maintenance = (getLastEvent() == HA_MAINTENANCE_START_EVT);
// It may be administratively disabled to handle partner's scope
// in case of failure. If this is the case we'll just handle our
// default scope (or no scope at all). The user will need to
// manually enable this server to handle partner's scope.
// If we're in the maintenance mode we serve all scopes because
// it is not a failover situation.
if (maintenance || config_->getThisServerConfig()->isAutoFailover()) {
query_filter_.serveFailoverScopes();
} else {
query_filter_.serveDefaultScopes();
}
adjustNetworkState();
communication_state_->clearRejectedLeaseUpdates();
// Log if the state machine is paused.
conditionalLogPausedState();
if (maintenance) {
// If we ended up in the partner-down state as a result of
// receiving the ha-maintenance-start command let's log it.
LOG_INFO(ha_logger, HA_MAINTENANCE_STARTED_IN_PARTNER_DOWN)
.arg(config_->getThisServerName());
}
} else if (getLastEvent() == HA_SYNCED_PARTNER_UNAVAILABLE_EVT) {
// Partner sent the ha-sync-complete-notify command to indicate that
// it has successfully synchronized its lease database but this server
// was unable to send heartbeat to this server. Enable the DHCP service
// and continue serving the clients in the partner-down state until the
// communication with the partner is fixed.
adjustNetworkState();
}
scheduleHeartbeat();
if (isMaintenanceCanceled() || isModelPaused()) {
postNextEvent(NOP_EVT);
return;
}
// Check if the clock skew is still acceptable. If not, transition to
// the terminated state.
if (shouldTerminate()) {
verboseTransition(HA_TERMINATED_ST);
return;
}
// Check if the partner state is valid per current configuration. If it is
// in an invalid state let's transition to the waiting state and stay there
// until the configuration is corrected.
if (isPartnerStateInvalid()) {
verboseTransition(HA_WAITING_ST);
return;
}
switch (communication_state_->getPartnerState()) {
case HA_COMMUNICATION_RECOVERY_ST:
case HA_PARTNER_DOWN_ST:
case HA_PARTNER_IN_MAINTENANCE_ST:
verboseTransition(HA_WAITING_ST);
break;
case HA_READY_ST:
// If partner allocated new leases for which it didn't send lease updates
// to us we should synchronize our database.
if (communication_state_->hasPartnerNewUnsentUpdates()) {
verboseTransition(HA_WAITING_ST);
} else {
// We did not miss any lease updates. There is no need to synchronize
// the database.
verboseTransition(getNormalState());
}
break;
case HA_TERMINATED_ST:
verboseTransition(HA_TERMINATED_ST);
break;
default:
postNextEvent(NOP_EVT);
}
}
void
HAService::partnerInMaintenanceStateHandler() {
// If we are transitioning from another state, we have to define new
// serving scopes appropriate for the new state. We don't do it if
// we remain in this state.
if (doOnEntry()) {
query_filter_.serveFailoverScopes();
adjustNetworkState();
// Log if the state machine is paused.
conditionalLogPausedState();
LOG_INFO(ha_logger, HA_MAINTENANCE_STARTED)
.arg(config_->getThisServerName());
}
scheduleHeartbeat();
if (isModelPaused()) {
postNextEvent(NOP_EVT);
return;
}
// Check if the clock skew is still acceptable. If not, transition to
// the terminated state.
if (shouldTerminate()) {
verboseTransition(HA_TERMINATED_ST);
return;
}
switch (communication_state_->getPartnerState()) {
case HA_UNAVAILABLE_ST:
verboseTransition(HA_PARTNER_DOWN_ST);
break;
default:
postNextEvent(NOP_EVT);
}
}
void
HAService::passiveBackupStateHandler() {
// If we are transitioning from another state, we have to define new
// serving scopes appropriate for the new state. We don't do it if
// we remain in this state.
if (doOnEntry()) {
query_filter_.serveDefaultScopes();
adjustNetworkState();
// In the passive-backup state we don't send heartbeat.
communication_state_->stopHeartbeat();
// Log if the state machine is paused.
conditionalLogPausedState();
}
postNextEvent(NOP_EVT);
}
void
HAService::readyStateHandler() {
// If we are transitioning from another state, we have to define new
// serving scopes appropriate for the new state. We don't do it if
// we remain in this state.
if (doOnEntry()) {
query_filter_.serveNoScopes();
adjustNetworkState();
communication_state_->clearRejectedLeaseUpdates();
// Log if the state machine is paused.
conditionalLogPausedState();
}
scheduleHeartbeat();
if (isMaintenanceCanceled() || isModelPaused()) {
postNextEvent(NOP_EVT);
return;
}
// Check if the clock skew is still acceptable. If not, transition to
// the terminated state.
if (shouldTerminate()) {
verboseTransition(HA_TERMINATED_ST);
return;
}
// Check if the partner state is valid per current configuration. If it is
// in an invalid state let's transition to the waiting state and stay there
// until the configuration is corrected.
if (isPartnerStateInvalid()) {
verboseTransition(HA_WAITING_ST);
return;
}
switch (communication_state_->getPartnerState()) {
case HA_HOT_STANDBY_ST:
case HA_LOAD_BALANCING_ST:
case HA_COMMUNICATION_RECOVERY_ST:
verboseTransition(getNormalState());
break;
case HA_IN_MAINTENANCE_ST:
verboseTransition(HA_PARTNER_IN_MAINTENANCE_ST);
break;
case HA_PARTNER_IN_MAINTENANCE_ST:
verboseTransition(HA_IN_MAINTENANCE_ST);
break;
case HA_READY_ST:
// If both servers are ready, the primary server "wins" and is
// transitioned first.
if (config_->getThisServerConfig()->getRole() == HAConfig::PeerConfig::PRIMARY) {
verboseTransition((config_->getHAMode() == HAConfig::LOAD_BALANCING ?
HA_LOAD_BALANCING_ST : HA_HOT_STANDBY_ST));
} else {
postNextEvent(NOP_EVT);
}
break;
case HA_TERMINATED_ST:
verboseTransition(HA_TERMINATED_ST);
break;
case HA_UNAVAILABLE_ST:
if (shouldPartnerDown()) {
verboseTransition(HA_PARTNER_DOWN_ST);
} else {
postNextEvent(NOP_EVT);
}
break;
default:
postNextEvent(NOP_EVT);
}
}
void
HAService::syncingStateHandler() {
// If we are transitioning from another state, we have to define new
// serving scopes appropriate for the new state. We don't do it if
// we remain in this state.
if (doOnEntry()) {
query_filter_.serveNoScopes();
adjustNetworkState();
communication_state_->clearRejectedLeaseUpdates();
// Log if the state machine is paused.
conditionalLogPausedState();
}
if (isMaintenanceCanceled() || isModelPaused()) {
postNextEvent(NOP_EVT);
return;
}
// Check if the clock skew is still acceptable. If not, transition to
// the terminated state.
if (shouldTerminate()) {
verboseTransition(HA_TERMINATED_ST);
return;
}
// Check if the partner state is valid per current configuration. If it is
// in an invalid state let's transition to the waiting state and stay there
// until the configuration is corrected.
if (isPartnerStateInvalid()) {
verboseTransition(HA_WAITING_ST);
return;
}
// We don't want to perform synchronous attempt to synchronize with
// a partner until we know that the partner is responding. Therefore,
// we wait for the heartbeat to complete successfully before we
// initiate the synchronization.
switch (communication_state_->getPartnerState()) {
case HA_TERMINATED_ST:
verboseTransition(HA_TERMINATED_ST);
return;
case HA_UNAVAILABLE_ST:
// If the partner appears to be offline, let's transition to the partner
// down state. Otherwise, we'd be stuck trying to synchronize with a
// dead partner.
if (shouldPartnerDown()) {
verboseTransition(HA_PARTNER_DOWN_ST);
} else {
postNextEvent(NOP_EVT);
}
break;
default:
// We don't want the heartbeat to interfere with the synchronization,
// so let's temporarily stop it.
communication_state_->stopHeartbeat();
// Timeout is configured in milliseconds. Need to convert to seconds.
unsigned int dhcp_disable_timeout =
static_cast<unsigned int>(config_->getSyncTimeout() / 1000);
if (dhcp_disable_timeout == 0) {
++dhcp_disable_timeout;
}
// Perform synchronous leases update.
std::string status_message;
int sync_status = synchronize(status_message,
config_->getFailoverPeerConfig(),
dhcp_disable_timeout);
// If the leases synchronization was successful, let's transition
// to the ready state.
if (sync_status == CONTROL_RESULT_SUCCESS) {
verboseTransition(HA_READY_ST);
} else {
// If the synchronization was unsuccessful we're back to the
// situation that the partner is unavailable and therefore
// we stay in the syncing state.
postNextEvent(NOP_EVT);
}
}
// Make sure that the heartbeat is re-enabled.
scheduleHeartbeat();
}
void
HAService::terminatedStateHandler() {
// If we are transitioning from another state, we have to define new
// serving scopes appropriate for the new state. We don't do it if
// we remain in this state.
if (doOnEntry()) {
query_filter_.serveDefaultScopes();
adjustNetworkState();
communication_state_->clearRejectedLeaseUpdates();
// In the terminated state we don't send heartbeat.
communication_state_->stopHeartbeat();
// Log if the state machine is paused.
conditionalLogPausedState();
LOG_ERROR(ha_logger, HA_TERMINATED)
.arg(config_->getThisServerName());
}
postNextEvent(NOP_EVT);
}
void
HAService::waitingStateHandler() {
// If we are transitioning from another state, we have to define new
// serving scopes appropriate for the new state. We don't do it if
// we remain in this state.
if (doOnEntry()) {
query_filter_.serveNoScopes();
adjustNetworkState();
communication_state_->clearRejectedLeaseUpdates();
// Log if the state machine is paused.
conditionalLogPausedState();
}
// Only schedule the heartbeat for non-backup servers.
if ((config_->getHAMode() != HAConfig::PASSIVE_BACKUP) &&
(config_->getThisServerConfig()->getRole() != HAConfig::PeerConfig::BACKUP)) {
scheduleHeartbeat();
}
if (isMaintenanceCanceled() || isModelPaused()) {
postNextEvent(NOP_EVT);
return;
}
// Backup server must remain in its own state.
if (config_->getThisServerConfig()->getRole() == HAConfig::PeerConfig::BACKUP) {
verboseTransition(HA_BACKUP_ST);
return;
}
// We're not a backup server, so we're either primary or secondary. If this is
// a passive-backup mode of operation, we're primary and we should transition
// to the passive-backup state.
if (config_->getHAMode() == HAConfig::PASSIVE_BACKUP) {
verboseTransition(HA_PASSIVE_BACKUP_ST);
return;
}
// Check if the clock skew is still acceptable. If not, transition to
// the terminated state.
if (shouldTerminate()) {
verboseTransition(HA_TERMINATED_ST);
return;
}
// Check if the partner state is valid per current configuration. If it is
// in an invalid state let's sit in the waiting state until the configuration
// is corrected.
if (isPartnerStateInvalid()) {
postNextEvent(NOP_EVT);
return;
}
switch (communication_state_->getPartnerState()) {
case HA_COMMUNICATION_RECOVERY_ST:
case HA_HOT_STANDBY_ST:
case HA_LOAD_BALANCING_ST:
case HA_IN_MAINTENANCE_ST:
case HA_PARTNER_DOWN_ST:
case HA_PARTNER_IN_MAINTENANCE_ST:
case HA_READY_ST:
// If we're configured to not synchronize lease database, proceed directly
// to the "ready" state.
verboseTransition(config_->amSyncingLeases() ? HA_SYNCING_ST : HA_READY_ST);
break;
case HA_SYNCING_ST:
postNextEvent(NOP_EVT);
break;
case HA_TERMINATED_ST: {
auto partner_in_terminated = communication_state_->getDurationSincePartnerStateTime();
if (!partner_in_terminated.is_not_a_date_time() &&
(partner_in_terminated.total_seconds()) / 60 >= HA_WAITING_TO_TERMINATED_ST_DELAY_MINUTES) {
LOG_WARN(ha_logger, HA_TERMINATED_PARTNER_DID_NOT_RESTART)
.arg(config_->getThisServerName())
.arg(HA_WAITING_TO_TERMINATED_ST_DELAY_MINUTES);
verboseTransition(HA_TERMINATED_ST);
break;
}
// We have checked above whether the clock skew is exceeding the threshold
// and we should terminate. If we're here, it means that the clock skew
// is acceptable. The partner may be still in the terminated state because
// it hasn't been restarted yet. Probably, this server is the first one
// being restarted after syncing the clocks. Let's just sit in the waiting
// state until the partner gets restarted.
LOG_INFO(ha_logger, HA_TERMINATED_RESTART_PARTNER)
.arg(config_->getThisServerName());
postNextEvent(NOP_EVT);
break;
}
case HA_WAITING_ST:
// If both servers are waiting, the primary server 'wins' and is
// transitioned to the next state first.
if (config_->getThisServerConfig()->getRole() == HAConfig::PeerConfig::PRIMARY) {
// If we're configured to not synchronize lease database, proceed directly
// to the "ready" state.
verboseTransition(config_->amSyncingLeases() ? HA_SYNCING_ST : HA_READY_ST);
} else {
postNextEvent(NOP_EVT);
}
break;
case HA_UNAVAILABLE_ST:
if (shouldPartnerDown()) {
verboseTransition(HA_PARTNER_DOWN_ST);
} else {
postNextEvent(NOP_EVT);
}
break;
default:
postNextEvent(NOP_EVT);
}
}
void
HAService::verboseTransition(const unsigned state) {
// Get current and new state name.
std::string current_state_name = getStateLabel(getCurrState());
std::string new_state_name = getStateLabel(state);
// Turn them to upper case so as they are better visible in the logs.
boost::to_upper(current_state_name);
boost::to_upper(new_state_name);
if (config_->getHAMode() != HAConfig::PASSIVE_BACKUP) {
// If this is load-balancing or hot-standby mode we also want to log
// partner's state.
auto partner_state = communication_state_->getPartnerState();
std::string partner_state_name = getStateLabel(partner_state);
boost::to_upper(partner_state_name);
// Log the transition.
LOG_INFO(ha_logger, HA_STATE_TRANSITION)
.arg(config_->getThisServerName())
.arg(current_state_name)
.arg(new_state_name)
.arg(partner_state_name);
} else {
// In the passive-backup mode we don't know the partner's state.
LOG_INFO(ha_logger, HA_STATE_TRANSITION_PASSIVE_BACKUP)
.arg(config_->getThisServerName())
.arg(current_state_name)
.arg(new_state_name);
}
// If we're transitioning directly from the "waiting" to "ready"
// state it indicates that the database synchronization is
// administratively disabled. Let's remind the user about this
// configuration setting.
if ((state == HA_READY_ST) && (getCurrState() == HA_WAITING_ST)) {
LOG_INFO(ha_logger, HA_CONFIG_LEASE_SYNCING_DISABLED_REMINDER)
.arg(config_->getThisServerName());
}
// Do the actual transition.
transition(state, getNextEvent());
// Inform the administrator whether or not lease updates are generated.
// Updates are never generated by a backup server so it doesn't make
// sense to log anything for the backup server.
if ((config_->getHAMode() != HAConfig::PASSIVE_BACKUP) &&
(config_->getThisServerConfig()->getRole() != HAConfig::PeerConfig::BACKUP)) {
if (shouldSendLeaseUpdates(config_->getFailoverPeerConfig())) {
LOG_INFO(ha_logger, HA_LEASE_UPDATES_ENABLED)
.arg(config_->getThisServerName())
.arg(new_state_name);
} else if (!config_->amSendingLeaseUpdates()) {
// Lease updates are administratively disabled.
LOG_INFO(ha_logger, HA_CONFIG_LEASE_UPDATES_DISABLED_REMINDER)
.arg(config_->getThisServerName())
.arg(new_state_name);
} else {
// Lease updates are not administratively disabled, but they
// are not issued because this is the backup server or because
// in this state the server should not generate lease updates.
LOG_INFO(ha_logger, HA_LEASE_UPDATES_DISABLED)
.arg(config_->getThisServerName())
.arg(new_state_name);
}
}
}
int
HAService::getNormalState() const {
if (config_->getThisServerConfig()->getRole() == HAConfig::PeerConfig::BACKUP) {
return (HA_BACKUP_ST);
}
switch (config_->getHAMode()) {
case HAConfig::LOAD_BALANCING:
return (HA_LOAD_BALANCING_ST);
case HAConfig::HOT_STANDBY:
return (HA_HOT_STANDBY_ST);
default:
return (HA_PASSIVE_BACKUP_ST);
}
}
bool
HAService::unpause() {
if (isModelPaused()) {
LOG_INFO(ha_logger, HA_STATE_MACHINE_CONTINUED)
.arg(config_->getThisServerName());
unpauseModel();
return (true);
}
return (false);
}
void
HAService::conditionalLogPausedState() const {
// Inform the administrator if the state machine is paused.
if (isModelPaused()) {
std::string state_name = stateToString(getCurrState());
boost::to_upper(state_name);
LOG_INFO(ha_logger, HA_STATE_MACHINE_PAUSED)
.arg(config_->getThisServerName())
.arg(state_name);
}
}
void
HAService::serveDefaultScopes() {
query_filter_.serveDefaultScopes();
}
void
HAService::serveFailoverScopes() {
query_filter_.serveFailoverScopes();
}
bool
HAService::inScope(dhcp::Pkt4Ptr& query4) {
return (inScopeInternal(query4));
}
bool
HAService::inScope(dhcp::Pkt6Ptr& query6) {
return (inScopeInternal(query6));
}
template<typename QueryPtrType>
bool
HAService::inScopeInternal(QueryPtrType& query) {
// Check if the query is in scope (should be processed by this server).
std::string scope_class;
const bool in_scope = query_filter_.inScope(query, scope_class);
// Whether or not the query is going to be processed by this server,
// we associate the query with the appropriate class.
query->addClass(dhcp::ClientClass(scope_class));
// The following is the part of the server failure detection algorithm.
// If the query should be processed by the partner we need to check if
// the partner responds. If the number of unanswered queries exceeds a
// configured threshold, we will consider the partner to be offline.
if (!in_scope && communication_state_->isCommunicationInterrupted()) {
communication_state_->analyzeMessage(query);
}
// Indicate if the query is in scope.
return (in_scope);
}
void
HAService::adjustNetworkState() {
std::string current_state_name = getStateLabel(getCurrState());
boost::to_upper(current_state_name);
// DHCP service should be enabled in the following states.
const bool should_enable = ((getCurrState() == HA_COMMUNICATION_RECOVERY_ST) ||
(getCurrState() == HA_LOAD_BALANCING_ST) ||
(getCurrState() == HA_HOT_STANDBY_ST) ||
(getCurrState() == HA_PARTNER_DOWN_ST) ||
(getCurrState() == HA_PARTNER_IN_MAINTENANCE_ST) ||
(getCurrState() == HA_PASSIVE_BACKUP_ST) ||
(getCurrState() == HA_TERMINATED_ST));
if (!should_enable && network_state_->isServiceEnabled()) {
current_state_name = getStateLabel(getCurrState());
boost::to_upper(current_state_name);
LOG_INFO(ha_logger, HA_LOCAL_DHCP_DISABLE)
.arg(config_->getThisServerName())
.arg(current_state_name);
network_state_->disableService(getLocalOrigin());
} else if (should_enable && !network_state_->isServiceEnabled()) {
current_state_name = getStateLabel(getCurrState());
boost::to_upper(current_state_name);
LOG_INFO(ha_logger, HA_LOCAL_DHCP_ENABLE)
.arg(config_->getThisServerName())
.arg(current_state_name);
network_state_->enableService(getLocalOrigin());
}
}
bool
HAService::shouldPartnerDown() const {
// Checking whether the communication with the partner is OK is the
// first step towards verifying if the server is up.
if (communication_state_->isCommunicationInterrupted()) {
// If the communication is interrupted, we also have to check
// whether the partner answers DHCP requests. The only cases
// when we don't (can't) do it are: the hot standby configuration
// in which this server is a primary and when the DHCP service is
// disabled so we can't analyze incoming traffic. Note that the
// primary server can't check delayed responses to the partner
// because the partner doesn't respond to any queries in this
// configuration.
if (network_state_->isServiceEnabled() &&
((config_->getHAMode() == HAConfig::LOAD_BALANCING) ||
(config_->getThisServerConfig()->getRole() == HAConfig::PeerConfig::STANDBY))) {
return (communication_state_->failureDetected());
}
// Hot standby / primary case.
return (true);
}
// Shouldn't transition to the partner down state.
return (false);
}
bool
HAService::shouldTerminate() const {
// Check if skew is fatally large.
bool should_terminate = communication_state_->clockSkewShouldTerminate();
// If not issue a warning if it's getting large.
if (!should_terminate) {
communication_state_->clockSkewShouldWarn();
// Check if we should terminate because the number of rejected leases
// has been exceeded.
should_terminate = communication_state_->rejectedLeaseUpdatesShouldTerminate();
}
return (should_terminate);
}
bool
HAService::isMaintenanceCanceled() const {
return (getLastEvent() == HA_MAINTENANCE_CANCEL_EVT);
}
bool
HAService::isPartnerStateInvalid() const {
switch (communication_state_->getPartnerState()) {
case HA_COMMUNICATION_RECOVERY_ST:
if (config_->getHAMode() != HAConfig::LOAD_BALANCING) {
LOG_WARN(ha_logger, HA_INVALID_PARTNER_STATE_COMMUNICATION_RECOVERY)
.arg(config_->getThisServerName());
return (true);
}
break;
case HA_HOT_STANDBY_ST:
if (config_->getHAMode() != HAConfig::HOT_STANDBY) {
LOG_WARN(ha_logger, HA_INVALID_PARTNER_STATE_HOT_STANDBY)
.arg(config_->getThisServerName());
return (true);
}
break;
case HA_LOAD_BALANCING_ST:
if (config_->getHAMode() != HAConfig::LOAD_BALANCING) {
LOG_WARN(ha_logger, HA_INVALID_PARTNER_STATE_LOAD_BALANCING)
.arg(config_->getThisServerName());
return (true);
}
break;
default:
;
}
return (false);
}
size_t
HAService::asyncSendLeaseUpdates(const dhcp::Pkt4Ptr& query,
const dhcp::Lease4CollectionPtr& leases,
const dhcp::Lease4CollectionPtr& deleted_leases,
const hooks::ParkingLotHandlePtr& parking_lot) {
// Get configurations of the peers. Exclude this instance.
HAConfig::PeerConfigMap peers_configs = config_->getOtherServersConfig();
size_t sent_num = 0;
// Schedule sending lease updates to each peer.
for (auto const& p : peers_configs) {
HAConfig::PeerConfigPtr conf = p.second;
// Check if the lease updates should be queued. This is the case when the
// server is in the communication-recovery state. Queued lease updates may
// be sent when the communication is re-established.
if (shouldQueueLeaseUpdates(conf)) {
// Lease updates for deleted leases.
for (auto const& l : *deleted_leases) {
// If a released lease is preserved in the database send the lease
// update to the partner. Otherwise, delete the lease.
if (l->state_ == Lease4::STATE_RELEASED) {
lease_update_backlog_.push(LeaseUpdateBacklog::ADD, l);
} else {
lease_update_backlog_.push(LeaseUpdateBacklog::DELETE, l);
}
}
// Lease updates for new allocations and updated leases.
for (auto const& l : *leases) {
lease_update_backlog_.push(LeaseUpdateBacklog::ADD, l);
}
continue;
}
// Check if the lease update should be sent to the server. If we're in
// the partner-down state we don't send lease updates to the partner.
if (!shouldSendLeaseUpdates(conf)) {
// If we decide to not send the lease updates to an active partner, we
// should make a record of it in the communication state. The partner
// can check if there were any unsent lease updates when he determines
// whether it should synchronize its database or not when it recovers
// from the partner-down state.
if (conf->getRole() != HAConfig::PeerConfig::BACKUP) {
communication_state_->increaseUnsentUpdateCount();
}
continue;
}
// Lease updates for deleted leases.
for (auto const& l : *deleted_leases) {
// If a released lease is preserved in the database send the lease
// update to the partner. Otherwise, delete the lease.
if (l->state_ == Lease4::STATE_RELEASED) {
asyncSendLeaseUpdate(query, conf, CommandCreator::createLease4Update(*l),
parking_lot);
} else {
asyncSendLeaseUpdate(query, conf, CommandCreator::createLease4Delete(*l),
parking_lot);
}
}
// Lease updates for new allocations and updated leases.
for (auto const& l : *leases) {
asyncSendLeaseUpdate(query, conf, CommandCreator::createLease4Update(*l),
parking_lot);
}
// If we're contacting a backup server from which we don't expect a
// response prior to responding to the DHCP client we don't count
// it.
if ((config_->amWaitingBackupAck() || (conf->getRole() != HAConfig::PeerConfig::BACKUP))) {
++sent_num;
}
}
return (sent_num);
}
size_t
HAService::asyncSendSingleLeaseUpdate(const dhcp::Pkt4Ptr& query,
const dhcp::Lease4Ptr& lease,
const hooks::ParkingLotHandlePtr& parking_lot) {
Lease4CollectionPtr leases(new Lease4Collection());
leases->push_back(lease);
Lease4CollectionPtr deleted_leases(new Lease4Collection());
return (asyncSendLeaseUpdates(query, leases, deleted_leases, parking_lot));
}
size_t
HAService::asyncSendLeaseUpdates(const dhcp::Pkt6Ptr& query,
const dhcp::Lease6CollectionPtr& leases,
const dhcp::Lease6CollectionPtr& deleted_leases,
const hooks::ParkingLotHandlePtr& parking_lot) {
// Get configurations of the peers. Exclude this instance.
HAConfig::PeerConfigMap peers_configs = config_->getOtherServersConfig();
size_t sent_num = 0;
// Schedule sending lease updates to each peer.
for (auto const& p : peers_configs) {
HAConfig::PeerConfigPtr conf = p.second;
// Check if the lease updates should be queued. This is the case when the
// server is in the communication-recovery state. Queued lease updates may
// be sent when the communication is re-established.
if (shouldQueueLeaseUpdates(conf)) {
for (auto const& l : *deleted_leases) {
// If a released lease is preserved in the database send the lease
// update to the partner. Otherwise, delete the lease.
if (l->state_ == Lease4::STATE_RELEASED) {
lease_update_backlog_.push(LeaseUpdateBacklog::ADD, l);
} else {
lease_update_backlog_.push(LeaseUpdateBacklog::DELETE, l);
}
}
// Lease updates for new allocations and updated leases.
for (auto const& l : *leases) {
lease_update_backlog_.push(LeaseUpdateBacklog::ADD, l);
}
continue;
}
// Check if the lease update should be sent to the server. If we're in
// the partner-down state we don't send lease updates to the partner.
if (!shouldSendLeaseUpdates(conf)) {
// If we decide to not send the lease updates to an active partner, we
// should make a record of it in the communication state. The partner
// can check if there were any unsent lease updates when he determines
// whether it should synchronize its database or not when it recovers
// from the partner-down state.
if (conf->getRole() != HAConfig::PeerConfig::BACKUP) {
communication_state_->increaseUnsentUpdateCount();
}
continue;
}
// If we're contacting a backup server from which we don't expect a
// response prior to responding to the DHCP client we don't count
// it.
if (config_->amWaitingBackupAck() || (conf->getRole() != HAConfig::PeerConfig::BACKUP)) {
++sent_num;
}
// Send new/updated leases and deleted leases in one command.
asyncSendLeaseUpdate(query, conf, CommandCreator::createLease6BulkApply(leases, deleted_leases),
parking_lot);
}
return (sent_num);
}
template<typename QueryPtrType>
bool
HAService::leaseUpdateComplete(QueryPtrType& query,
const ParkingLotHandlePtr& parking_lot) {
if (MultiThreadingMgr::instance().getMode()) {
std::lock_guard<std::mutex> lock(mutex_);
return (leaseUpdateCompleteInternal(query, parking_lot));
} else {
return (leaseUpdateCompleteInternal(query, parking_lot));
}
}
template<typename QueryPtrType>
bool
HAService::leaseUpdateCompleteInternal(QueryPtrType& query,
const ParkingLotHandlePtr& parking_lot) {
auto it = pending_requests_.find(query);
// If there are no more pending requests for this query, let's unpark
// the DHCP packet.
if (it == pending_requests_.end() || (--pending_requests_[query] <= 0)) {
if (parking_lot) {
parking_lot->unpark(query);
}
// If we have unparked the packet we can clear pending requests for
// this query.
if (it != pending_requests_.end()) {
pending_requests_.erase(it);
}
return (true);
}
return (false);
}
template<typename QueryPtrType>
void
HAService::updatePendingRequest(QueryPtrType& query) {
if (MultiThreadingMgr::instance().getMode()) {
std::lock_guard<std::mutex> lock(mutex_);
updatePendingRequestInternal(query);
} else {
updatePendingRequestInternal(query);
}
}
template<typename QueryPtrType>
void
HAService::updatePendingRequestInternal(QueryPtrType& query) {
if (pending_requests_.count(query) == 0) {
pending_requests_[query] = 1;
} else {
++pending_requests_[query];
}
}
template<typename QueryPtrType>
void
HAService::asyncSendLeaseUpdate(const QueryPtrType& query,
const HAConfig::PeerConfigPtr& config,
const ConstElementPtr& command,
const ParkingLotHandlePtr& parking_lot) {
// Create HTTP/1.1 request including our command.
PostHttpRequestJsonPtr request = boost::make_shared<PostHttpRequestJson>
(HttpRequest::Method::HTTP_POST, "/", HttpVersion::HTTP_11(),
HostHttpHeader(config->getUrl().getStrippedHostname()));
config->addBasicAuthHttpHeader(request);
request->setBodyAsJson(command);
request->finalize();
// Response object should also be created because the HTTP client needs
// to know the type of the expected response.
HttpResponseJsonPtr response = boost::make_shared<HttpResponseJson>();
// When possible we prefer to pass weak pointers to the queries, rather
// than shared pointers, to avoid memory leaks in case cross reference
// between the pointers.
boost::weak_ptr<typename QueryPtrType::element_type> weak_query(query);
// Schedule asynchronous HTTP request.
client_->asyncSendRequest(config->getUrl(), config->getTlsContext(),
request, response,
[this, weak_query, parking_lot, config]
(const boost::system::error_code& ec,
const HttpResponsePtr& response,
const std::string& error_str) {
// Get the shared pointer of the query. The server should keep the
// pointer to the query and then park it. Therefore, we don't really
// expect it to be null. If it is null, something is really wrong.
QueryPtrType query = weak_query.lock();
if (!query) {
isc_throw(Unexpected, "query is null while receiving response from"
" HA peer. This is programmatic error");
}
// There are four possible groups of errors during the lease update.
// One is the IO error causing issues in communication with the peer.
// Another one is an HTTP parsing error. The third type occurs when
// the partner receives the command but it is invalid or there is
// an internal processing error. Finally, the forth type is when the
// conflict status code is returned in the response indicating that
// the lease update does not match the partner's configuration.
bool lease_update_success = true;
bool lease_update_conflict = false;
// Handle first two groups of errors.
if (ec || !error_str.empty()) {
LOG_WARN(ha_logger, HA_LEASE_UPDATE_COMMUNICATIONS_FAILED)
.arg(config_->getThisServerName())
.arg(query->getLabel())
.arg(config->getLogLabel())
.arg(ec ? ec.message() : error_str);
// Communication error, so let's drop parked packet. The DHCP
// response will not be sent.
lease_update_success = false;
} else {
try {
int rcode = 0;
auto args = verifyAsyncResponse(response, rcode);
// In the v6 case the server may return a list of failed lease
// updates and we should log them.
logFailedLeaseUpdates(query, args);
} catch (const ConflictError& ex) {
// Handle forth group of errors.
lease_update_conflict = true;
lease_update_success = false;
communication_state_->reportRejectedLeaseUpdate(query);
LOG_WARN(ha_logger, HA_LEASE_UPDATE_CONFLICT)
.arg(config_->getThisServerName())
.arg(query->getLabel())
.arg(config->getLogLabel())
.arg(ex.what());
} catch (const std::exception& ex) {
// Handle third group of errors.
LOG_WARN(ha_logger, HA_LEASE_UPDATE_FAILED)
.arg(config_->getThisServerName())
.arg(query->getLabel())
.arg(config->getLogLabel())
.arg(ex.what());
// Error while doing an update. The DHCP response will not be sent.
lease_update_success = false;
}
}
// We don't care about the result of the lease update to the backup server.
// It is a best effort update.
if (config->getRole() != HAConfig::PeerConfig::BACKUP) {
// If the lease update was unsuccessful we may need to set the partner
// state as unavailable.
if (!lease_update_success) {
// Do not set it as unavailable if it was a conflict because the
// partner actually responded.
if (!lease_update_conflict) {
// If we were unable to communicate with the partner we set partner's
// state as unavailable.
communication_state_->setPartnerUnavailable();
}
} else {
// Lease update successful and we may need to clear some previously
// rejected lease updates.
communication_state_->reportSuccessfulLeaseUpdate(query);
}
}
// It is possible to configure the server to not wait for a response from
// the backup server before we unpark the packet and respond to the client.
// Here we check if we're dealing with such situation.
if (config_->amWaitingBackupAck() || (config->getRole() != HAConfig::PeerConfig::BACKUP)) {
// We're expecting a response from the backup server or it is not
// a backup server and the lease update was unsuccessful. In such
// case the DHCP exchange fails.
if (!lease_update_success) {
if (parking_lot) {
parking_lot->drop(query);
}
}
} else {
// This was a response from the backup server and we're configured to
// not wait for their acknowledgments, so there is nothing more to do.
return;
}
if (leaseUpdateComplete(query, parking_lot)) {
// If we have finished sending the lease updates we need to run the
// state machine until the state machine finds that additional events
// are required, such as next heartbeat or a lease update. The runModel()
// may transition to another state, schedule asynchronous tasks etc.
// Then it returns control to the DHCP server.
runModel(HA_LEASE_UPDATES_COMPLETE_EVT);
}
},
HttpClient::RequestTimeout(TIMEOUT_DEFAULT_HTTP_CLIENT_REQUEST),
std::bind(&HAService::clientConnectHandler, this, ph::_1, ph::_2),
std::bind(&HAService::clientHandshakeHandler, this, ph::_1),
std::bind(&HAService::clientCloseHandler, this, ph::_1)
);
// The number of pending requests is the number of requests for which we
// expect an acknowledgment prior to responding to the DHCP clients. If
// we're configured to wait for the acks from the backups or it is not
// a backup increase the number of pending requests.
if (config_->amWaitingBackupAck() || (config->getRole() != HAConfig::PeerConfig::BACKUP)) {
// Request scheduled, so update the request counters for the query.
updatePendingRequest(query);
}
}
bool
HAService::shouldSendLeaseUpdates(const HAConfig::PeerConfigPtr& peer_config) const {
// Never send lease updates if they are administratively disabled.
if (!config_->amSendingLeaseUpdates()) {
return (false);
}
// Always send updates to the backup server.
if (peer_config->getRole() == HAConfig::PeerConfig::BACKUP) {
return (true);
}
// Never send updates if this is a backup server.
if (config_->getThisServerConfig()->getRole() == HAConfig::PeerConfig::BACKUP) {
return (false);
}
// In other case, whether we send lease updates or not depends on our
// state.
switch (getCurrState()) {
case HA_HOT_STANDBY_ST:
case HA_LOAD_BALANCING_ST:
case HA_PARTNER_IN_MAINTENANCE_ST:
return (true);
default:
;
}
return (false);
}
bool
HAService::shouldQueueLeaseUpdates(const HAConfig::PeerConfigPtr& peer_config) const {
if (!config_->amSendingLeaseUpdates()) {
return (false);
}
if (peer_config->getRole() == HAConfig::PeerConfig::BACKUP) {
return (false);
}
return (getCurrState() == HA_COMMUNICATION_RECOVERY_ST);
}
void
HAService::logFailedLeaseUpdates(const PktPtr& query,
const ConstElementPtr& args) const {
// If there are no arguments, it means that the update was successful.
if (!args || (args->getType() != Element::map)) {
return;
}
// Instead of duplicating the code between the failed-deleted-leases and
// failed-leases, let's just have one function that does it for both.
auto log_proc = [](const PktPtr query, const ConstElementPtr& args,
const std::string& param_name, const log::MessageID& mesid) {
// Check if there are any failed leases.
auto failed_leases = args->get(param_name);
// The failed leases must be a list.
if (failed_leases && (failed_leases->getType() == Element::list)) {
// Go over the failed leases and log each of them.
for (int i = 0; i < failed_leases->size(); ++i) {
auto lease = failed_leases->get(i);
if (lease->getType() == Element::map) {
// ip-address
auto ip_address = lease->get("ip-address");
// lease type
auto lease_type = lease->get("type");
// error-message
auto error_message = lease->get("error-message");
LOG_INFO(ha_logger, mesid)
.arg(query->getLabel())
.arg(lease_type && (lease_type->getType() == Element::string) ?
lease_type->stringValue() : "(unknown)")
.arg(ip_address && (ip_address->getType() == Element::string) ?
ip_address->stringValue() : "(unknown)")
.arg(error_message && (error_message->getType() == Element::string) ?
error_message->stringValue() : "(unknown)");
}
}
}
};
// Process "failed-deleted-leases"
log_proc(query, args, "failed-deleted-leases", HA_LEASE_UPDATE_DELETE_FAILED_ON_PEER);
// Process "failed-leases".
log_proc(query, args, "failed-leases", HA_LEASE_UPDATE_CREATE_UPDATE_FAILED_ON_PEER);
}
ConstElementPtr
HAService::processStatusGet() const {
ElementPtr ha_servers = Element::createMap();
// Local part
ElementPtr local = Element::createMap();
HAConfig::PeerConfig::Role role;
role = config_->getThisServerConfig()->getRole();
std::string role_txt = HAConfig::PeerConfig::roleToString(role);
local->set("role", Element::create(role_txt));
int state = getCurrState();
try {
local->set("state", Element::create(stateToString(state)));
} catch (...) {
// Empty string on error.
local->set("state", Element::create(std::string()));
}
std::set<std::string> scopes = query_filter_.getServedScopes();
ElementPtr list = Element::createList();
for (auto const& scope : scopes) {
list->add(Element::create(scope));
}
local->set("scopes", list);
local->set("server-name", Element::create(config_->getThisServerName()));
auto const my_time(communication_state_->getMyTimeAtSkew());
if (my_time.is_not_a_date_time()) {
local->set("system-time", Element::create());
} else {
local->set("system-time", Element::create(ptimeToText(my_time, 0)));
}
ha_servers->set("local", local);
// Do not include remote server information if this is a backup server or
// we're in the passive-backup mode.
if ((config_->getHAMode() == HAConfig::PASSIVE_BACKUP) ||
(config_->getThisServerConfig()->getRole() == HAConfig::PeerConfig::BACKUP)) {
return (ha_servers);
}
// Remote part
ElementPtr remote = communication_state_->getReport();
try {
role = config_->getFailoverPeerConfig()->getRole();
role_txt = HAConfig::PeerConfig::roleToString(role);
remote->set("role", Element::create(role_txt));
} catch (...) {
remote->set("role", Element::create(std::string()));
}
remote->set("server-name", Element::create(config_->getFailoverPeerConfig()->getName()));
ha_servers->set("remote", remote);
return (ha_servers);
}
ConstElementPtr
HAService::processHeartbeat() {
ElementPtr arguments = Element::createMap();
std::string state_label = getState(getCurrState())->getLabel();
arguments->set("state", Element::create(state_label));
std::string date_time = HttpDateTime().rfc1123Format();
arguments->set("date-time", Element::create(date_time));
auto scopes = query_filter_.getServedScopes();
ElementPtr scopes_list = Element::createList();
for (auto const& scope : scopes) {
scopes_list->add(Element::create(scope));
}
arguments->set("scopes", scopes_list);
arguments->set("unsent-update-count",
Element::create(static_cast<int64_t>(communication_state_->getUnsentUpdateCount())));
return (createAnswer(CONTROL_RESULT_SUCCESS, "HA peer status returned.",
arguments));
}
ConstElementPtr
HAService::processHAReset() {
if (getCurrState() == HA_WAITING_ST) {
return (createAnswer(CONTROL_RESULT_SUCCESS, "HA state machine already in WAITING state."));
}
verboseTransition(HA_WAITING_ST);
runModel(NOP_EVT);
return (createAnswer(CONTROL_RESULT_SUCCESS, "HA state machine reset."));
}
void
HAService::asyncSendHeartbeat() {
HAConfig::PeerConfigPtr partner_config = config_->getFailoverPeerConfig();
// If the sync_complete_notified_ is true it means that the partner
// notified us that it had completed lease database synchronization.
// We confirm that the partner is operational by sending the heartbeat
// to it. Regardless if the partner responds to our heartbeats or not,
// we should clear this flag. But, since we need the current value in
// the async call handler, we save it in the local variable before
// clearing it.
bool sync_complete_notified = sync_complete_notified_;
sync_complete_notified_ = false;
// Create HTTP/1.1 request including our command.
PostHttpRequestJsonPtr request = boost::make_shared<PostHttpRequestJson>
(HttpRequest::Method::HTTP_POST, "/", HttpVersion::HTTP_11(),
HostHttpHeader(partner_config->getUrl().getStrippedHostname()));
partner_config->addBasicAuthHttpHeader(request);
request->setBodyAsJson(CommandCreator::createHeartbeat(config_->getThisServerName(),
server_type_));
request->finalize();
// Response object should also be created because the HTTP client needs
// to know the type of the expected response.
HttpResponseJsonPtr response = boost::make_shared<HttpResponseJson>();
// Schedule asynchronous HTTP request.
client_->asyncSendRequest(partner_config->getUrl(),
partner_config->getTlsContext(),
request, response,
[this, partner_config, sync_complete_notified]
(const boost::system::error_code& ec,
const HttpResponsePtr& response,
const std::string& error_str) {
// There are three possible groups of errors during the heartbeat.
// One is the IO error causing issues in communication with the peer.
// Another one is an HTTP parsing error. The last type of error is
// when non-success error code is returned in the response carried
// in the HTTP message or if the JSON response is otherwise broken.
bool heartbeat_success = true;
// Handle first two groups of errors.
if (ec || !error_str.empty()) {
LOG_WARN(ha_logger, HA_HEARTBEAT_COMMUNICATIONS_FAILED)
.arg(config_->getThisServerName())
.arg(partner_config->getLogLabel())
.arg(ec ? ec.message() : error_str);
heartbeat_success = false;
} else {
// Handle third group of errors.
try {
// Response must contain arguments and the arguments must
// be a map.
int rcode = 0;
ConstElementPtr args = verifyAsyncResponse(response, rcode);
if (!args || args->getType() != Element::map) {
isc_throw(CtrlChannelError, "returned arguments in the response"
" must be a map");
}
// Response must include partner's state.
ConstElementPtr state = args->get("state");
if (!state || state->getType() != Element::string) {
isc_throw(CtrlChannelError, "server state not returned in response"
" to a ha-heartbeat command or it is not a string");
}
// Remember the partner's state. This may throw if the returned
// state is invalid.
communication_state_->setPartnerState(state->stringValue());
ConstElementPtr date_time = args->get("date-time");
if (!date_time || date_time->getType() != Element::string) {
isc_throw(CtrlChannelError, "date-time not returned in response"
" to a ha-heartbeat command or it is not a string");
}
// Note the time returned by the partner to calculate the clock skew.
communication_state_->setPartnerTime(date_time->stringValue());
// Remember the scopes served by the partner.
try {
auto scopes = args->get("scopes");
communication_state_->setPartnerScopes(scopes);
} catch (...) {
// We don't want to fail if the scopes are missing because
// this would be incompatible with old HA hook library
// versions. We may make it mandatory one day, but during
// upgrades of existing HA setup it would be a real issue
// if we failed here.
}
// unsent-update-count was not present in earlier HA versions.
// Let's check if the partner has sent the parameter. We initialized
// the counter to 0, and it remains 0 if the partner doesn't send it.
// It effectively means that we don't track partner's unsent updates
// as in the earlier HA versions.
auto unsent_update_count = args->get("unsent-update-count");
if (unsent_update_count) {
if (unsent_update_count->getType() != Element::integer) {
isc_throw(CtrlChannelError, "unsent-update-count returned in"
" the ha-heartbeat response is not an integer");
}
communication_state_->setPartnerUnsentUpdateCount(static_cast<uint64_t>
(unsent_update_count->intValue()));
}
} catch (const std::exception& ex) {
LOG_WARN(ha_logger, HA_HEARTBEAT_FAILED)
.arg(config_->getThisServerName())
.arg(partner_config->getLogLabel())
.arg(ex.what());
heartbeat_success = false;
}
}
// If heartbeat was successful, let's mark the connection with the
// peer as healthy.
if (heartbeat_success) {
communication_state_->poke();
} else {
// We were unable to retrieve partner's state, so let's mark it
// as unavailable.
communication_state_->setPartnerUnavailable();
// Log if the communication is interrupted.
if (communication_state_->isCommunicationInterrupted()) {
LOG_WARN(ha_logger, HA_COMMUNICATION_INTERRUPTED)
.arg(config_->getThisServerName())
.arg(partner_config->getName());
}
}
startHeartbeat();
// Even though the partner notified us about the synchronization completion,
// we still can't communicate with the partner. Let's continue serving
// the clients until the link is fixed.
if (sync_complete_notified && !heartbeat_success) {
postNextEvent(HA_SYNCED_PARTNER_UNAVAILABLE_EVT);
}
// Whatever the result of the heartbeat was, the state machine needs
// to react to this. Let's run the state machine until the state machine
// finds that some new events are required, i.e. next heartbeat or
// lease update. The runModel() may transition to another state, schedule
// asynchronous tasks etc. Then it returns control to the DHCP server.
runModel(HA_HEARTBEAT_COMPLETE_EVT);
},
HttpClient::RequestTimeout(TIMEOUT_DEFAULT_HTTP_CLIENT_REQUEST),
std::bind(&HAService::clientConnectHandler, this, ph::_1, ph::_2),
std::bind(&HAService::clientHandshakeHandler, this, ph::_1),
std::bind(&HAService::clientCloseHandler, this, ph::_1)
);
}
void
HAService::scheduleHeartbeat() {
if (!communication_state_->isHeartbeatRunning()) {
startHeartbeat();
}
}
void
HAService::startHeartbeat() {
if (config_->getHeartbeatDelay() > 0) {
communication_state_->startHeartbeat(config_->getHeartbeatDelay(),
std::bind(&HAService::asyncSendHeartbeat,
this));
}
}
void
HAService::asyncDisableDHCPService(HttpClient& http_client,
const HAConfig::PeerConfigPtr& remote_config,
const unsigned int max_period,
PostRequestCallback post_request_action) {
// Create HTTP/1.1 request including our command.
PostHttpRequestJsonPtr request = boost::make_shared<PostHttpRequestJson>
(HttpRequest::Method::HTTP_POST, "/", HttpVersion::HTTP_11(),
HostHttpHeader(remote_config->getUrl().getStrippedHostname()));
remote_config->addBasicAuthHttpHeader(request);
request->setBodyAsJson(CommandCreator::createDHCPDisable(getRemoteOrigin(),
max_period,
server_type_));
request->finalize();
// Response object should also be created because the HTTP client needs
// to know the type of the expected response.
HttpResponseJsonPtr response = boost::make_shared<HttpResponseJson>();
// Schedule asynchronous HTTP request.
http_client.asyncSendRequest(remote_config->getUrl(),
remote_config->getTlsContext(),
request, response,
[this, remote_config, post_request_action]
(const boost::system::error_code& ec,
const HttpResponsePtr& response,
const std::string& error_str) {
// There are three possible groups of errors during the heartbeat.
// One is the IO error causing issues in communication with the peer.
// Another one is an HTTP parsing error. The last type of error is
// when non-success error code is returned in the response carried
// in the HTTP message or if the JSON response is otherwise broken.
int rcode = 0;
std::string error_message;
// Handle first two groups of errors.
if (ec || !error_str.empty()) {
error_message = (ec ? ec.message() : error_str);
LOG_ERROR(ha_logger, HA_DHCP_DISABLE_COMMUNICATIONS_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
} else {
// Handle third group of errors.
try {
static_cast<void>(verifyAsyncResponse(response, rcode));
} catch (const std::exception& ex) {
error_message = ex.what();
LOG_ERROR(ha_logger, HA_DHCP_DISABLE_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
}
}
// If there was an error communicating with the partner, mark the
// partner as unavailable.
if (!error_message.empty()) {
communication_state_->setPartnerUnavailable();
}
// Invoke post request action if it was specified.
if (post_request_action) {
post_request_action(error_message.empty(),
error_message,
rcode);
}
},
HttpClient::RequestTimeout(TIMEOUT_DEFAULT_HTTP_CLIENT_REQUEST),
std::bind(&HAService::clientConnectHandler, this, ph::_1, ph::_2),
std::bind(&HAService::clientHandshakeHandler, this, ph::_1),
std::bind(&HAService::clientCloseHandler, this, ph::_1)
);
}
void
HAService::asyncEnableDHCPService(HttpClient& http_client,
const HAConfig::PeerConfigPtr& remote_config,
PostRequestCallback post_request_action) {
// Create HTTP/1.1 request including our command.
PostHttpRequestJsonPtr request = boost::make_shared<PostHttpRequestJson>
(HttpRequest::Method::HTTP_POST, "/", HttpVersion::HTTP_11(),
HostHttpHeader(remote_config->getUrl().getStrippedHostname()));
remote_config->addBasicAuthHttpHeader(request);
request->setBodyAsJson(CommandCreator::createDHCPEnable(getRemoteOrigin(),
server_type_));
request->finalize();
// Response object should also be created because the HTTP client needs
// to know the type of the expected response.
HttpResponseJsonPtr response = boost::make_shared<HttpResponseJson>();
// Schedule asynchronous HTTP request.
http_client.asyncSendRequest(remote_config->getUrl(),
remote_config->getTlsContext(),
request, response,
[this, remote_config, post_request_action]
(const boost::system::error_code& ec,
const HttpResponsePtr& response,
const std::string& error_str) {
// There are three possible groups of errors during the heartbeat.
// One is the IO error causing issues in communication with the peer.
// Another one is an HTTP parsing error. The last type of error is
// when non-success error code is returned in the response carried
// in the HTTP message or if the JSON response is otherwise broken.
int rcode = 0;
std::string error_message;
// Handle first two groups of errors.
if (ec || !error_str.empty()) {
error_message = (ec ? ec.message() : error_str);
LOG_ERROR(ha_logger, HA_DHCP_ENABLE_COMMUNICATIONS_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
} else {
// Handle third group of errors.
try {
static_cast<void>(verifyAsyncResponse(response, rcode));
} catch (const std::exception& ex) {
error_message = ex.what();
LOG_ERROR(ha_logger, HA_DHCP_ENABLE_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
}
}
// If there was an error communicating with the partner, mark the
// partner as unavailable.
if (!error_message.empty()) {
communication_state_->setPartnerUnavailable();
}
// Invoke post request action if it was specified.
if (post_request_action) {
post_request_action(error_message.empty(),
error_message,
rcode);
}
},
HttpClient::RequestTimeout(TIMEOUT_DEFAULT_HTTP_CLIENT_REQUEST),
std::bind(&HAService::clientConnectHandler, this, ph::_1, ph::_2),
std::bind(&HAService::clientHandshakeHandler, this, ph::_1),
std::bind(&HAService::clientCloseHandler, this, ph::_1)
);
}
void
HAService::localDisableDHCPService() {
network_state_->disableService(getLocalOrigin());
}
void
HAService::localEnableDHCPService() {
network_state_->enableService(getLocalOrigin());
}
void
HAService::asyncSyncLeases() {
PostSyncCallback null_action;
// Timeout is configured in milliseconds. Need to convert to seconds.
unsigned int dhcp_disable_timeout =
static_cast<unsigned int>(config_->getSyncTimeout() / 1000);
if (dhcp_disable_timeout == 0) {
// Ensure that we always use at least 1 second timeout.
dhcp_disable_timeout = 1;
}
lease_sync_filter_.apply();
asyncSyncLeases(*client_, config_->getFailoverPeerConfig(),
dhcp_disable_timeout, LeasePtr(), null_action);
}
void
HAService::asyncSyncLeases(http::HttpClient& http_client,
const HAConfig::PeerConfigPtr& remote_config,
const unsigned int max_period,
const dhcp::LeasePtr& last_lease,
PostSyncCallback post_sync_action,
const bool dhcp_disabled) {
// Synchronization starts with a command to disable DHCP service of the
// peer from which we're fetching leases. We don't want the other server
// to allocate new leases while we fetch from it. The DHCP service will
// be disabled for a certain amount of time and will be automatically
// re-enabled if we die during the synchronization.
asyncDisableDHCPService(http_client, remote_config, max_period,
[this, &http_client, remote_config, max_period, last_lease,
post_sync_action, dhcp_disabled]
(const bool success, const std::string& error_message, const int) {
// If we have successfully disabled the DHCP service on the peer,
// we can start fetching the leases.
if (success) {
// The last argument indicates that disabling the DHCP
// service on the partner server was successful.
asyncSyncLeasesInternal(http_client, remote_config, max_period,
last_lease, post_sync_action, true);
} else {
post_sync_action(success, error_message, dhcp_disabled);
}
});
}
void
HAService::asyncSyncLeasesInternal(http::HttpClient& http_client,
const HAConfig::PeerConfigPtr& remote_config,
const unsigned int max_period,
const dhcp::LeasePtr& last_lease,
PostSyncCallback post_sync_action,
const bool dhcp_disabled) {
// Create HTTP/1.1 request including our command.
PostHttpRequestJsonPtr request = boost::make_shared<PostHttpRequestJson>
(HttpRequest::Method::HTTP_POST, "/", HttpVersion::HTTP_11(),
HostHttpHeader(remote_config->getUrl().getStrippedHostname()));
remote_config->addBasicAuthHttpHeader(request);
if (server_type_ == HAServerType::DHCPv4) {
request->setBodyAsJson(CommandCreator::createLease4GetPage(
boost::dynamic_pointer_cast<Lease4>(last_lease), config_->getSyncPageLimit()));
} else {
request->setBodyAsJson(CommandCreator::createLease6GetPage(
boost::dynamic_pointer_cast<Lease6>(last_lease), config_->getSyncPageLimit()));
}
request->finalize();
// Response object should also be created because the HTTP client needs
// to know the type of the expected response.
HttpResponseJsonPtr response = boost::make_shared<HttpResponseJson>();
// Schedule asynchronous HTTP request.
http_client.asyncSendRequest(remote_config->getUrl(),
remote_config->getTlsContext(),
request, response,
[this, remote_config, post_sync_action, &http_client, max_period, dhcp_disabled]
(const boost::system::error_code& ec,
const HttpResponsePtr& response,
const std::string& error_str) {
// Holds last lease received on the page of leases. If the last
// page was hit, this value remains null.
LeasePtr last_lease;
// There are three possible groups of errors during the heartbeat.
// One is the IO error causing issues in communication with the peer.
// Another one is an HTTP parsing error. The last type of error is
// when non-success error code is returned in the response carried
// in the HTTP message or if the JSON response is otherwise broken.
std::string error_message;
// Handle first two groups of errors.
if (ec || !error_str.empty()) {
error_message = (ec ? ec.message() : error_str);
LOG_ERROR(ha_logger, HA_LEASES_SYNC_COMMUNICATIONS_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
} else {
// Handle third group of errors.
try {
int rcode = 0;
ConstElementPtr args = verifyAsyncResponse(response, rcode);
// Arguments must be a map.
if (args && (args->getType() != Element::map)) {
isc_throw(CtrlChannelError,
"arguments in the received response must be a map");
}
ConstElementPtr leases = args->get("leases");
if (!leases || (leases->getType() != Element::list)) {
isc_throw(CtrlChannelError,
"server response does not contain leases argument or this"
" argument is not a list");
}
// Iterate over the leases and update the database as appropriate.
auto const& leases_element = leases->listValue();
LOG_INFO(ha_logger, HA_LEASES_SYNC_LEASE_PAGE_RECEIVED)
.arg(config_->getThisServerName())
.arg(leases_element.size())
.arg(remote_config->getLogLabel());
// Count actually applied leases.
uint64_t applied_lease_count = 0;
for (auto l = leases_element.begin(); l != leases_element.end(); ++l) {
try {
if (server_type_ == HAServerType::DHCPv4) {
Lease4Ptr lease = Lease4::fromElement(*l);
// If we're not on the last page and we're processing final lease on
// this page, let's record the lease as input to the next
// lease4-get-page command.
if ((leases_element.size() >= config_->getSyncPageLimit()) &&
(l + 1 == leases_element.end())) {
last_lease = boost::dynamic_pointer_cast<Lease>(lease);
}
if (!lease_sync_filter_.shouldSync(lease)) {
continue;
}
// Check if there is such lease in the database already.
Lease4Ptr existing_lease = LeaseMgrFactory::instance().getLease4(lease->addr_);
if (!existing_lease) {
// There is no such lease, so let's add it.
LeaseMgrFactory::instance().addLease(lease);
++applied_lease_count;
} else if (existing_lease->cltt_ < lease->cltt_) {
// If the existing lease is older than the fetched lease, update
// the lease in our local database.
// Update lease current expiration time with value received from the
// database. Some database backends reject operations on the lease if
// the current expiration time value does not match what is stored.
Lease::syncCurrentExpirationTime(*existing_lease, *lease);
LeaseMgrFactory::instance().updateLease4(lease);
++applied_lease_count;
} else {
LOG_DEBUG(ha_logger, DBGLVL_TRACE_BASIC, HA_LEASE_SYNC_STALE_LEASE4_SKIP)
.arg(config_->getThisServerName())
.arg(lease->addr_.toText())
.arg(lease->subnet_id_);
}
} else {
Lease6Ptr lease = Lease6::fromElement(*l);
// If we're not on the last page and we're processing final lease on
// this page, let's record the lease as input to the next
// lease6-get-page command.
if ((leases_element.size() >= config_->getSyncPageLimit()) &&
(l + 1 == leases_element.end())) {
last_lease = boost::dynamic_pointer_cast<Lease>(lease);
}
if (!lease_sync_filter_.shouldSync(lease)) {
continue;
}
// Check if there is such lease in the database already.
Lease6Ptr existing_lease = LeaseMgrFactory::instance().getLease6(lease->type_,
lease->addr_);
if (!existing_lease) {
// There is no such lease, so let's add it.
LeaseMgrFactory::instance().addLease(lease);
++applied_lease_count;
} else if (existing_lease->cltt_ < lease->cltt_) {
// If the existing lease is older than the fetched lease, update
// the lease in our local database.
// Update lease current expiration time with value received from the
// database. Some database backends reject operations on the lease if
// the current expiration time value does not match what is stored.
Lease::syncCurrentExpirationTime(*existing_lease, *lease);
LeaseMgrFactory::instance().updateLease6(lease);
++applied_lease_count;
} else {
LOG_DEBUG(ha_logger, DBGLVL_TRACE_BASIC, HA_LEASE_SYNC_STALE_LEASE6_SKIP)
.arg(config_->getThisServerName())
.arg(lease->addr_.toText())
.arg(lease->subnet_id_);
}
}
} catch (const std::exception& ex) {
LOG_WARN(ha_logger, HA_LEASE_SYNC_FAILED)
.arg(config_->getThisServerName())
.arg((*l)->str())
.arg(ex.what());
}
}
LOG_INFO(ha_logger, HA_LEASES_SYNC_APPLIED_LEASES)
.arg(config_->getThisServerName())
.arg(applied_lease_count);
} catch (const std::exception& ex) {
error_message = ex.what();
LOG_ERROR(ha_logger, HA_LEASES_SYNC_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
}
}
// If there was an error communicating with the partner, mark the
// partner as unavailable.
if (!error_message.empty()) {
communication_state_->setPartnerUnavailable();
} else if (last_lease) {
// This indicates that there are more leases to be fetched.
// Therefore, we have to send another leaseX-get-page command.
asyncSyncLeases(http_client, remote_config, max_period, last_lease,
post_sync_action, dhcp_disabled);
return;
}
// Invoke post synchronization action if it was specified.
if (post_sync_action) {
post_sync_action(error_message.empty(),
error_message,
dhcp_disabled);
}
},
HttpClient::RequestTimeout(config_->getSyncTimeout()),
std::bind(&HAService::clientConnectHandler, this, ph::_1, ph::_2),
std::bind(&HAService::clientHandshakeHandler, this, ph::_1),
std::bind(&HAService::clientCloseHandler, this, ph::_1)
);
}
ConstElementPtr
HAService::processSynchronize(const std::string& server_name,
const unsigned int max_period) {
HAConfig::PeerConfigPtr remote_config;
try {
remote_config = config_->getPeerConfig(server_name);
} catch (const std::exception& ex) {
return (createAnswer(CONTROL_RESULT_ERROR, ex.what()));
}
// We must not synchronize with self.
if (remote_config->getName() == config_->getThisServerName()) {
return (createAnswer(CONTROL_RESULT_ERROR, "'" + remote_config->getName()
+ "' points to local server but should point to a partner"));
}
std::string answer_message;
int sync_status = synchronize(answer_message, remote_config, max_period);
return (createAnswer(sync_status, answer_message));
}
int
HAService::synchronize(std::string& status_message,
const HAConfig::PeerConfigPtr& remote_config,
const unsigned int max_period) {
lease_sync_filter_.apply();
IOServicePtr io_service(new IOService());
HttpClient client(io_service, false);
asyncSyncLeases(client, remote_config, max_period, Lease4Ptr(),
[&](const bool success, const std::string& error_message,
const bool dhcp_disabled) {
// If there was a fatal error while fetching the leases, let's
// log an error message so as it can be included in the response
// to the controlling client.
if (!success) {
status_message = error_message;
}
// Whether or not there was an error while fetching the leases,
// we need to re-enable the DHCP service on the peer if the
// DHCP service was disabled in the course of synchronization.
if (dhcp_disabled) {
// If the synchronization was completed successfully let's
// try to send the ha-sync-complete-notify command to the
// partner.
if (success) {
asyncSyncCompleteNotify(client, remote_config,
[&](const bool success,
const std::string& error_message,
const int rcode) {
// This command may not be supported by the partner when it
// runs an older Kea version. In that case, send the dhcp-enable
// command as in previous Kea version.
if (rcode == CONTROL_RESULT_COMMAND_UNSUPPORTED) {
asyncEnableDHCPService(client, remote_config,
[&](const bool success,
const std::string& error_message,
const int) {
// It is possible that we have already recorded an error
// message while synchronizing the lease database. Don't
// override the existing error message.
if (!success && status_message.empty()) {
status_message = error_message;
}
// The synchronization process is completed, so let's break
// the IO service so as we can return the response to the
// controlling client.
io_service->stop();
});
} else {
// ha-sync-complete-notify command was delivered to the partner.
// The synchronization process ends here.
if (!success && status_message.empty()) {
status_message = error_message;
}
io_service->stop();
}
});
} else {
// Synchronization was unsuccessful. Send the dhcp-enable command to
// re-enable the DHCP service. Note, that we don't send the
// ha-sync-complete-notify command in this case. It is only sent in
// the case when synchronization ends successfully.
asyncEnableDHCPService(client, remote_config,
[&](const bool success,
const std::string& error_message,
const int) {
if (!success && status_message.empty()) {
status_message = error_message;
}
// The synchronization process is completed, so let's break
// the IO service so as we can return the response to the
// controlling client.
io_service->stop();
});
}
} else {
// Also stop IO service if there is no need to enable DHCP
// service.
io_service->stop();
}
});
LOG_INFO(ha_logger, HA_SYNC_START)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel());
// Measure duration of the synchronization.
Stopwatch stopwatch;
// Run the IO service until it is stopped by any of the callbacks. This
// makes it synchronous.
io_service->run();
// End measuring duration.
stopwatch.stop();
client.stop();
io_service->stopAndPoll();
// If an error message has been recorded, return an error to the controlling
// client.
if (!status_message.empty()) {
postNextEvent(HA_SYNCING_FAILED_EVT);
LOG_ERROR(ha_logger, HA_SYNC_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(status_message);
return (CONTROL_RESULT_ERROR);
}
// Everything was fine, so let's return a success.
status_message = "Lease database synchronization complete.";
postNextEvent(HA_SYNCING_SUCCEEDED_EVT);
LOG_INFO(ha_logger, HA_SYNC_SUCCESSFUL)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(stopwatch.logFormatLastDuration());
return (CONTROL_RESULT_SUCCESS);
}
void
HAService::asyncSendLeaseUpdatesFromBacklog(HttpClient& http_client,
const HAConfig::PeerConfigPtr& config,
PostRequestCallback post_request_action) {
if (lease_update_backlog_.size() == 0) {
post_request_action(true, "", CONTROL_RESULT_SUCCESS);
return;
}
ConstElementPtr command;
if (server_type_ == HAServerType::DHCPv4) {
LeaseUpdateBacklog::OpType op_type;
Lease4Ptr lease = boost::dynamic_pointer_cast<Lease4>(lease_update_backlog_.pop(op_type));
if (op_type == LeaseUpdateBacklog::ADD) {
command = CommandCreator::createLease4Update(*lease);
} else {
command = CommandCreator::createLease4Delete(*lease);
}
} else {
command = CommandCreator::createLease6BulkApply(lease_update_backlog_);
}
// Create HTTP/1.1 request including our command.
PostHttpRequestJsonPtr request = boost::make_shared<PostHttpRequestJson>
(HttpRequest::Method::HTTP_POST, "/", HttpVersion::HTTP_11(),
HostHttpHeader(config->getUrl().getStrippedHostname()));
config->addBasicAuthHttpHeader(request);
request->setBodyAsJson(command);
request->finalize();
// Response object should also be created because the HTTP client needs
// to know the type of the expected response.
HttpResponseJsonPtr response = boost::make_shared<HttpResponseJson>();
http_client.asyncSendRequest(config->getUrl(), config->getTlsContext(),
request, response,
[this, &http_client, config, post_request_action]
(const boost::system::error_code& ec,
const HttpResponsePtr& response,
const std::string& error_str) {
int rcode = 0;
std::string error_message;
if (ec || !error_str.empty()) {
error_message = (ec ? ec.message() : error_str);
LOG_WARN(ha_logger, HA_LEASES_BACKLOG_COMMUNICATIONS_FAILED)
.arg(config_->getThisServerName())
.arg(config->getLogLabel())
.arg(ec ? ec.message() : error_str);
} else {
// Handle third group of errors.
try {
auto args = verifyAsyncResponse(response, rcode);
} catch (const std::exception& ex) {
error_message = ex.what();
LOG_WARN(ha_logger, HA_LEASES_BACKLOG_FAILED)
.arg(config_->getThisServerName())
.arg(config->getLogLabel())
.arg(ex.what());
}
}
// Recursively send all outstanding lease updates or break when an
// error occurs. In DHCPv6, this is a single iteration because we use
// lease6-bulk-apply, which combines many lease updates in a single
// transaction. In the case of DHCPv4, each update is sent in its own
// transaction.
if (error_message.empty()) {
asyncSendLeaseUpdatesFromBacklog(http_client, config, post_request_action);
} else {
post_request_action(error_message.empty(), error_message, rcode);
}
});
}
bool
HAService::sendLeaseUpdatesFromBacklog() {
auto num_updates = lease_update_backlog_.size();
if (num_updates == 0) {
LOG_INFO(ha_logger, HA_LEASES_BACKLOG_NOTHING_TO_SEND)
.arg(config_->getThisServerName());
return (true);
}
IOServicePtr io_service(new IOService());
HttpClient client(io_service, false);
auto remote_config = config_->getFailoverPeerConfig();
bool updates_successful = true;
LOG_INFO(ha_logger, HA_LEASES_BACKLOG_START)
.arg(config_->getThisServerName())
.arg(num_updates)
.arg(remote_config->getName());
asyncSendLeaseUpdatesFromBacklog(client, remote_config,
[&](const bool success, const std::string&, const int) {
io_service->stop();
updates_successful = success;
});
// Measure duration of the updates.
Stopwatch stopwatch;
// Run the IO service until it is stopped by the callback. This makes it synchronous.
io_service->run();
// End measuring duration.
stopwatch.stop();
client.stop();
io_service->stopAndPoll();
if (updates_successful) {
LOG_INFO(ha_logger, HA_LEASES_BACKLOG_SUCCESS)
.arg(config_->getThisServerName())
.arg(remote_config->getName())
.arg(stopwatch.logFormatLastDuration());
}
return (updates_successful);
}
void
HAService::asyncSendHAReset(HttpClient& http_client,
const HAConfig::PeerConfigPtr& config,
PostRequestCallback post_request_action) {
ConstElementPtr command = CommandCreator::createHAReset(config_->getThisServerName(),
server_type_);
// Create HTTP/1.1 request including our command.
PostHttpRequestJsonPtr request = boost::make_shared<PostHttpRequestJson>
(HttpRequest::Method::HTTP_POST, "/", HttpVersion::HTTP_11(),
HostHttpHeader(config->getUrl().getStrippedHostname()));
config->addBasicAuthHttpHeader(request);
request->setBodyAsJson(command);
request->finalize();
// Response object should also be created because the HTTP client needs
// to know the type of the expected response.
HttpResponseJsonPtr response = boost::make_shared<HttpResponseJson>();
http_client.asyncSendRequest(config->getUrl(), config->getTlsContext(),
request, response,
[this, config, post_request_action]
(const boost::system::error_code& ec,
const HttpResponsePtr& response,
const std::string& error_str) {
int rcode = 0;
std::string error_message;
if (ec || !error_str.empty()) {
error_message = (ec ? ec.message() : error_str);
LOG_WARN(ha_logger, HA_RESET_COMMUNICATIONS_FAILED)
.arg(config_->getThisServerName())
.arg(config->getLogLabel())
.arg(ec ? ec.message() : error_str);
} else {
// Handle third group of errors.
try {
auto args = verifyAsyncResponse(response, rcode);
} catch (const std::exception& ex) {
error_message = ex.what();
LOG_WARN(ha_logger, HA_RESET_FAILED)
.arg(config_->getThisServerName())
.arg(config->getLogLabel())
.arg(ex.what());
}
}
post_request_action(error_message.empty(), error_message, rcode);
});
}
bool
HAService::sendHAReset() {
IOServicePtr io_service(new IOService());
HttpClient client(io_service, false);
auto remote_config = config_->getFailoverPeerConfig();
bool reset_successful = true;
asyncSendHAReset(client, remote_config,
[&](const bool success, const std::string&, const int) {
io_service->stop();
reset_successful = success;
});
// Run the IO service until it is stopped by the callback. This makes it synchronous.
io_service->run();
client.stop();
io_service->stopAndPoll();
return (reset_successful);
}
ConstElementPtr
HAService::processScopes(const std::vector<std::string>& scopes) {
try {
query_filter_.serveScopes(scopes);
adjustNetworkState();
} catch (const std::exception& ex) {
return (createAnswer(CONTROL_RESULT_ERROR, ex.what()));
}
return (createAnswer(CONTROL_RESULT_SUCCESS, "New HA scopes configured."));
}
ConstElementPtr
HAService::processContinue() {
if (unpause()) {
return (createAnswer(CONTROL_RESULT_SUCCESS, "HA state machine continues."));
}
return (createAnswer(CONTROL_RESULT_SUCCESS, "HA state machine is not paused."));
}
ConstElementPtr
HAService::processMaintenanceNotify(const bool cancel, const std::string& state) {
if (cancel) {
if (getCurrState() != HA_IN_MAINTENANCE_ST) {
return (createAnswer(CONTROL_RESULT_ERROR, "Unable to cancel the"
" maintenance for the server not in the"
" in-maintenance state."));
}
try {
communication_state_->setPartnerState(state);
} catch (...) {
// Hopefully the received state is correct. If it isn't, let's set the
// partner state to unavailable and count on the state machine to resolve.
communication_state_->setPartnerUnavailable();
}
postNextEvent(HA_MAINTENANCE_CANCEL_EVT);
// In rare cases the previous state may be the server's current state. Transitioning
// to it would cause a deadlock and the server will remain stuck in maintenance.
// In these cases let's simply transition to the waiting state and the state machine
// should solve it.
verboseTransition(getPrevState() == HA_IN_MAINTENANCE_ST ? HA_WAITING_ST : getPrevState());
runModel(NOP_EVT);
// Communicate the new state to the partner.
ElementPtr arguments = Element::createMap();
std::string state_label = getState(getCurrState())->getLabel();
arguments->set("state", Element::create(state_label));
return (createAnswer(CONTROL_RESULT_SUCCESS, "Server maintenance canceled.", arguments));
}
switch (getCurrState()) {
case HA_BACKUP_ST:
case HA_PARTNER_IN_MAINTENANCE_ST:
case HA_TERMINATED_ST:
// The reason why we don't return an error result here is that we have to
// have a way to distinguish between the errors caused by the communication
// issues and the cases when there is no communication error but the server
// is not allowed to enter the in-maintenance state. In the former case, the
// partner would go to partner-down. In the case signaled by the special
// result code entering the maintenance state is not allowed.
return (createAnswer(HA_CONTROL_RESULT_MAINTENANCE_NOT_ALLOWED,
"Unable to transition the server from the "
+ stateToString(getCurrState()) + " to"
" in-maintenance state."));
default:
verboseTransition(HA_IN_MAINTENANCE_ST);
runModel(HA_MAINTENANCE_NOTIFY_EVT);
}
return (createAnswer(CONTROL_RESULT_SUCCESS, "Server is in-maintenance state."));
}
ConstElementPtr
HAService::processMaintenanceStart() {
switch (getCurrState()) {
case HA_BACKUP_ST:
case HA_IN_MAINTENANCE_ST:
case HA_PARTNER_IN_MAINTENANCE_ST:
case HA_TERMINATED_ST:
return (createAnswer(CONTROL_RESULT_ERROR, "Unable to transition the server from"
" the " + stateToString(getCurrState()) + " to"
" partner-in-maintenance state."));
default:
;
}
HAConfig::PeerConfigPtr remote_config = config_->getFailoverPeerConfig();
// Create HTTP/1.1 request including ha-maintenance-notify command
// with the cancel flag set to false.
PostHttpRequestJsonPtr request = boost::make_shared<PostHttpRequestJson>
(HttpRequest::Method::HTTP_POST, "/", HttpVersion::HTTP_11(),
HostHttpHeader(remote_config->getUrl().getStrippedHostname()));
remote_config->addBasicAuthHttpHeader(request);
request->setBodyAsJson(CommandCreator::createMaintenanceNotify(config_->getThisServerName(),
false, getCurrState(), server_type_));
request->finalize();
// Response object should also be created because the HTTP client needs
// to know the type of the expected response.
HttpResponseJsonPtr response = boost::make_shared<HttpResponseJson>();
IOServicePtr io_service(new IOService());
HttpClient client(io_service, false);
boost::system::error_code captured_ec;
std::string captured_error_message;
int captured_rcode = 0;
// Schedule asynchronous HTTP request.
client.asyncSendRequest(remote_config->getUrl(),
remote_config->getTlsContext(),
request, response,
[this, remote_config, &io_service, &captured_ec, &captured_error_message,
&captured_rcode]
(const boost::system::error_code& ec,
const HttpResponsePtr& response,
const std::string& error_str) {
io_service->stop();
// There are three possible groups of errors. One is the IO error
// causing issues in communication with the peer. Another one is
// an HTTP parsing error. The last type of error is when non-success
// error code is returned in the response carried in the HTTP message
// or if the JSON response is otherwise broken.
std::string error_message;
// Handle first two groups of errors.
if (ec || !error_str.empty()) {
error_message = (ec ? ec.message() : error_str);
LOG_ERROR(ha_logger, HA_MAINTENANCE_NOTIFY_COMMUNICATIONS_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
} else {
// Handle third group of errors.
try {
static_cast<void>(verifyAsyncResponse(response, captured_rcode));
} catch (const std::exception& ex) {
error_message = ex.what();
LOG_ERROR(ha_logger, HA_MAINTENANCE_NOTIFY_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
}
}
// If there was an error communicating with the partner, mark the
// partner as unavailable.
if (!error_message.empty()) {
communication_state_->setPartnerUnavailable();
}
captured_ec = ec;
captured_error_message = error_message;
},
HttpClient::RequestTimeout(TIMEOUT_DEFAULT_HTTP_CLIENT_REQUEST),
std::bind(&HAService::clientConnectHandler, this, ph::_1, ph::_2),
std::bind(&HAService::clientHandshakeHandler, this, ph::_1),
std::bind(&HAService::clientCloseHandler, this, ph::_1)
);
// Run the IO service until it is stopped by any of the callbacks. This
// makes it synchronous.
io_service->run();
client.stop();
io_service->stopAndPoll();
// If there was a communication problem with the partner we assume that
// the partner is already down while we receive this command.
if (captured_ec || (captured_rcode == CONTROL_RESULT_ERROR)) {
postNextEvent(HA_MAINTENANCE_START_EVT);
verboseTransition(HA_PARTNER_DOWN_ST);
runModel(NOP_EVT);
return (createAnswer(CONTROL_RESULT_SUCCESS,
"Server is now in the partner-down state as its"
" partner appears to be offline for maintenance."));
} else if (captured_rcode == CONTROL_RESULT_SUCCESS) {
// If the partner responded indicating no error it means that the
// partner has been transitioned to the in-maintenance state. In that
// case we transition to the partner-in-maintenance state.
postNextEvent(HA_MAINTENANCE_START_EVT);
verboseTransition(HA_PARTNER_IN_MAINTENANCE_ST);
runModel(NOP_EVT);
} else {
// Partner server returned a special status code which means that it can't
// transition to the partner-in-maintenance state.
return (createAnswer(CONTROL_RESULT_ERROR, "Unable to transition to the"
" partner-in-maintenance state. The partner server responded"
" with the following message to the ha-maintenance-notify"
" command: " + captured_error_message + "."));
}
return (createAnswer(CONTROL_RESULT_SUCCESS,
"Server is now in the partner-in-maintenance state"
" and its partner is in-maintenance state. The partner"
" can be now safely shut down."));
}
ConstElementPtr
HAService::processMaintenanceCancel() {
if (getCurrState() != HA_PARTNER_IN_MAINTENANCE_ST) {
return (createAnswer(CONTROL_RESULT_ERROR, "Unable to cancel maintenance"
" request because the server is not in the"
" partner-in-maintenance state."));
}
// This is the state the server will transition to if the notification to the
// partner is successful.
int next_state = getPrevState() == HA_PARTNER_IN_MAINTENANCE_ST ? HA_WAITING_ST : getPrevState();
HAConfig::PeerConfigPtr remote_config = config_->getFailoverPeerConfig();
// Create HTTP/1.1 request including ha-maintenance-notify command
// with the cancel flag set to true.
PostHttpRequestJsonPtr request = boost::make_shared<PostHttpRequestJson>
(HttpRequest::Method::HTTP_POST, "/", HttpVersion::HTTP_11(),
HostHttpHeader(remote_config->getUrl().getStrippedHostname()));
remote_config->addBasicAuthHttpHeader(request);
request->setBodyAsJson(CommandCreator::createMaintenanceNotify(config_->getThisServerName(),
true,
next_state,
server_type_));
request->finalize();
// Response object should also be created because the HTTP client needs
// to know the type of the expected response.
HttpResponseJsonPtr response = boost::make_shared<HttpResponseJson>();
IOServicePtr io_service(new IOService());
HttpClient client(io_service, false);
std::string error_message;
// Schedule asynchronous HTTP request.
client.asyncSendRequest(remote_config->getUrl(),
remote_config->getTlsContext(),
request, response,
[this, remote_config, &io_service, &error_message]
(const boost::system::error_code& ec,
const HttpResponsePtr& response,
const std::string& error_str) {
io_service->stop();
// Handle first two groups of errors.
if (ec || !error_str.empty()) {
error_message = (ec ? ec.message() : error_str);
LOG_ERROR(ha_logger, HA_MAINTENANCE_NOTIFY_CANCEL_COMMUNICATIONS_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
} else {
// Handle third group of errors.
try {
int rcode = 0;
ConstElementPtr args = verifyAsyncResponse(response, rcode);
// Partner's state has changed after the notification. However, we don't know
// its new state. We'll check if the partner returned its state. If it didn't,
// we set the unavailable state as a default.
communication_state_->setPartnerUnavailable();
// Newer Kea versions return the state of the notified server.
// Older versions don't, so the arguments may not be present.
if (args && args->getType() == Element::map) {
// Arguments may include partner's state.
ConstElementPtr state = args->get("state");
if (state) {
if (state->getType() != Element::string) {
isc_throw(CtrlChannelError, "server state not returned in response"
" to a ha-heartbeat command or it is not a string");
}
communication_state_->setPartnerState(state->stringValue());
}
}
} catch (const std::exception& ex) {
error_message = ex.what();
LOG_ERROR(ha_logger, HA_MAINTENANCE_NOTIFY_CANCEL_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
}
}
// If there was an error communicating with the partner, mark the
// partner as unavailable.
if (!error_message.empty()) {
communication_state_->setPartnerUnavailable();
}
},
HttpClient::RequestTimeout(TIMEOUT_DEFAULT_HTTP_CLIENT_REQUEST),
std::bind(&HAService::clientConnectHandler, this, ph::_1, ph::_2),
std::bind(&HAService::clientHandshakeHandler, this, ph::_1),
std::bind(&HAService::clientCloseHandler, this, ph::_1)
);
// Run the IO service until it is stopped by any of the callbacks. This
// makes it synchronous.
io_service->run();
client.stop();
io_service->stopAndPoll();
// There was an error in communication with the partner or the
// partner was unable to revert its state.
if (!error_message.empty()) {
return (createAnswer(CONTROL_RESULT_ERROR,
"Unable to cancel maintenance. The partner server responded"
" with the following message to the ha-maintenance-notify"
" command: " + error_message + "."));
}
// Successfully reverted partner's state. Let's also revert our state to the
// previous one. Avoid returning to the partner-in-maintenance if it was
// the previous state.
postNextEvent(HA_MAINTENANCE_CANCEL_EVT);
verboseTransition(next_state);
runModel(NOP_EVT);
return (createAnswer(CONTROL_RESULT_SUCCESS,
"Server maintenance successfully canceled."));
}
void
HAService::asyncSyncCompleteNotify(HttpClient& http_client,
const HAConfig::PeerConfigPtr& remote_config,
PostRequestCallback post_request_action) {
// Create HTTP/1.1 request including our command.
PostHttpRequestJsonPtr request = boost::make_shared<PostHttpRequestJson>
(HttpRequest::Method::HTTP_POST, "/", HttpVersion::HTTP_11(),
HostHttpHeader(remote_config->getUrl().getStrippedHostname()));
remote_config->addBasicAuthHttpHeader(request);
request->setBodyAsJson(CommandCreator::createSyncCompleteNotify(getRemoteOrigin(),
config_->getThisServerName(),
server_type_));
request->finalize();
// Response object should also be created because the HTTP client needs
// to know the type of the expected response.
HttpResponseJsonPtr response = boost::make_shared<HttpResponseJson>();
// Schedule asynchronous HTTP request.
http_client.asyncSendRequest(remote_config->getUrl(),
remote_config->getTlsContext(),
request, response,
[this, remote_config, post_request_action]
(const boost::system::error_code& ec,
const HttpResponsePtr& response,
const std::string& error_str) {
// There are three possible groups of errors. One is the IO error
// causing issues in communication with the peer. Another one is an
// HTTP parsing error. The last type of error is when non-success
// error code is returned in the response carried in the HTTP message
// or if the JSON response is otherwise broken.
int rcode = 0;
std::string error_message;
// Handle first two groups of errors.
if (ec || !error_str.empty()) {
error_message = (ec ? ec.message() : error_str);
LOG_ERROR(ha_logger, HA_SYNC_COMPLETE_NOTIFY_COMMUNICATIONS_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
} else {
// Handle third group of errors.
try {
static_cast<void>(verifyAsyncResponse(response, rcode));
} catch (const CommandUnsupportedError& ex) {
rcode = CONTROL_RESULT_COMMAND_UNSUPPORTED;
} catch (const std::exception& ex) {
error_message = ex.what();
LOG_ERROR(ha_logger, HA_SYNC_COMPLETE_NOTIFY_FAILED)
.arg(config_->getThisServerName())
.arg(remote_config->getLogLabel())
.arg(error_message);
}
}
// If there was an error communicating with the partner, mark the
// partner as unavailable.
if (!error_message.empty()) {
communication_state_->setPartnerUnavailable();
}
// Invoke post request action if it was specified.
if (post_request_action) {
post_request_action(error_message.empty(),
error_message,
rcode);
}
},
HttpClient::RequestTimeout(TIMEOUT_DEFAULT_HTTP_CLIENT_REQUEST),
std::bind(&HAService::clientConnectHandler, this, ph::_1, ph::_2),
std::bind(&HAService::clientHandshakeHandler, this, ph::_1),
std::bind(&HAService::clientCloseHandler, this, ph::_1)
);
}
ConstElementPtr
HAService::processSyncCompleteNotify(const unsigned int origin_id) {
if (getCurrState() == HA_PARTNER_DOWN_ST) {
sync_complete_notified_ = true;
// We're in the partner-down state and the partner notified us
// that it has synchronized its database. We can't enable the
// service yet, because it may result in some new lease allocations
// that the partner would miss (we don't send lease updates in the
// partner-down state). We must first send the heartbeat and let
// the state machine resolve the situation between the partners.
// It may unblock the network service.
network_state_->disableService(getLocalOrigin());
}
// Release the network state lock for the remote origin because we have
// acquired the local network state lock above (partner-down state), or
// we don't need the lock (other states).
network_state_->enableService(origin_id);
return (createAnswer(CONTROL_RESULT_SUCCESS,
"Server successfully notified about the synchronization completion."));
}
ConstElementPtr
HAService::verifyAsyncResponse(const HttpResponsePtr& response, int& rcode) {
// Set the return code to error in case of early throw.
rcode = CONTROL_RESULT_ERROR;
// The response must cast to JSON type.
HttpResponseJsonPtr json_response =
boost::dynamic_pointer_cast<HttpResponseJson>(response);
if (!json_response) {
isc_throw(CtrlChannelError, "no valid HTTP response found");
}
// Body holds the response to our command.
ConstElementPtr body = json_response->getBodyAsJson();
if (!body) {
isc_throw(CtrlChannelError, "no body found in the response");
}
// Body should contain a list of responses from multiple servers.
if (body->getType() != Element::list) {
// Some control agent errors are returned as a map.
if (body->getType() == Element::map) {
ElementPtr list = Element::createList();
ElementPtr answer = Element::createMap();
answer->set(CONTROL_RESULT, Element::create(rcode));
ConstElementPtr text = body->get(CONTROL_TEXT);
if (text) {
answer->set(CONTROL_TEXT, text);
}
list->add(answer);
body = list;
} else {
isc_throw(CtrlChannelError, "body of the response must be a list");
}
}
// There must be at least one response.
if (body->empty()) {
isc_throw(CtrlChannelError, "list of responses must not be empty");
}
// Check if the status code of the first response. We don't support multiple
// at this time, because we always send a request to a single location.
ConstElementPtr args = parseAnswer(rcode, body->get(0));
if (rcode == CONTROL_RESULT_SUCCESS) {
return (args);
}
std::ostringstream s;
// The empty status can occur for the lease6-bulk-apply command. In that
// case, the response may contain conflicted or erred leases within the
// arguments, rather than globally. For other error cases let's construct
// the error message from the global values.
if (rcode != CONTROL_RESULT_EMPTY) {
// Include an error text if available.
if (args && args->getType() == Element::string) {
s << args->stringValue() << " (";
}
// Include an error code.
s << "error code " << rcode << ")";
}
switch (rcode) {
case CONTROL_RESULT_COMMAND_UNSUPPORTED:
isc_throw(CommandUnsupportedError, s.str());
case CONTROL_RESULT_CONFLICT:
isc_throw(ConflictError, s.str());
case CONTROL_RESULT_EMPTY:
// Handle the lease6-bulk-apply error cases.
if (args && (args->getType() == Element::map)) {
auto failed_leases = args->get("failed-leases");
if (!failed_leases || (failed_leases->getType() != Element::list)) {
// If there are no failed leases there is nothing to do.
break;
}
auto conflict = false;
ConstElementPtr conflict_error_message;
for (auto i = 0; i < failed_leases->size(); ++i) {
auto lease = failed_leases->get(i);
if (!lease || lease->getType() != Element::map) {
continue;
}
auto result = lease->get("result");
if (!result || result->getType() != Element::integer) {
continue;
}
auto error_message = lease->get("error-message");
// Error status code takes precedence over the conflict.
if (result->intValue() == CONTROL_RESULT_ERROR) {
if (error_message && error_message->getType()) {
s << error_message->stringValue() << " (";
}
s << "error code " << result->intValue() << ")";
isc_throw(CtrlChannelError, s.str());
}
if (result->intValue() == CONTROL_RESULT_CONFLICT) {
// Let's record the conflict but there may still be some
// leases with an error status code, so do not throw the
// conflict exception yet.
conflict = true;
conflict_error_message = error_message;
}
}
if (conflict) {
// There are no errors. There are only conflicts. Throw
// appropriate exception.
if (conflict_error_message &&
(conflict_error_message->getType() == Element::string)) {
s << conflict_error_message->stringValue() << " (";
}
s << "error code " << CONTROL_RESULT_CONFLICT << ")";
isc_throw(ConflictError, s.str());
}
}
break;
default:
isc_throw(CtrlChannelError, s.str());
}
return (args);
}
bool
HAService::clientConnectHandler(const boost::system::error_code& ec, int tcp_native_fd) {
// If client is running it's own IOService we do NOT want to
// register the socket with IfaceMgr.
if (client_->getThreadIOService()) {
return (true);
}
// If things look OK register the socket with Interface Manager. Note
// we don't register if the FD is < 0 to avoid an exception throw.
// It is unlikely that this will occur but we want to be liberal
// and avoid issues.
if ((!ec || (ec.value() == boost::asio::error::in_progress))
&& (tcp_native_fd >= 0)) {
// External socket callback is a NOP. Ready events handlers are
// run by an explicit call IOService ready in kea-dhcp<n> code.
// We are registering the socket only to interrupt main-thread
// select().
IfaceMgr::instance().addExternalSocket(tcp_native_fd,
std::bind(&HAService::socketReadyHandler, this, ph::_1)
);
}
// If ec.value() == boost::asio::error::already_connected, we should already
// be registered, so nothing to do. If it is any other value, then connect
// failed and Connection logic should handle that, not us, so no matter
// what happens we're returning true.
return (true);
}
void
HAService::socketReadyHandler(int tcp_native_fd) {
// If the socket is ready but does not belong to one of our client's
// ongoing transactions, we close it. This will unregister it from
// IfaceMgr and ensure the client starts over with a fresh connection
// if it needs to do so.
client_->closeIfOutOfBand(tcp_native_fd);
}
void
HAService::clientCloseHandler(int tcp_native_fd) {
if (tcp_native_fd >= 0) {
IfaceMgr::instance().deleteExternalSocket(tcp_native_fd);
}
}
size_t
HAService::pendingRequestSize() {
if (MultiThreadingMgr::instance().getMode()) {
std::lock_guard<std::mutex> lock(mutex_);
return (pending_requests_.size());
} else {
return (pending_requests_.size());
}
}
template<typename QueryPtrType>
int
HAService::getPendingRequest(const QueryPtrType& query) {
if (MultiThreadingMgr::instance().getMode()) {
std::lock_guard<std::mutex> lock(mutex_);
return (getPendingRequestInternal(query));
} else {
return (getPendingRequestInternal(query));
}
}
template<typename QueryPtrType>
int
HAService::getPendingRequestInternal(const QueryPtrType& query) {
if (pending_requests_.count(query) == 0) {
return (0);
} else {
return (pending_requests_[query]);
}
}
void
HAService::checkPermissionsClientAndListener() {
// Since this function is used as CS callback all exceptions must be
// suppressed (except the @ref MultiThreadingInvalidOperation), unlikely
// though they may be.
// The @ref MultiThreadingInvalidOperation is propagated to the scope of the
// @ref MultiThreadingCriticalSection constructor.
try {
if (client_) {
client_->checkPermissions();
}
if (listener_) {
listener_->checkPermissions();
}
} catch (const isc::MultiThreadingInvalidOperation& ex) {
LOG_ERROR(ha_logger, HA_PAUSE_CLIENT_LISTENER_ILLEGAL)
.arg(config_->getThisServerName())
.arg(ex.what());
// The exception needs to be propagated to the caller of the
// @ref MultiThreadingCriticalSection constructor.
throw;
} catch (const std::exception& ex) {
LOG_ERROR(ha_logger, HA_PAUSE_CLIENT_LISTENER_FAILED)
.arg(config_->getThisServerName())
.arg(ex.what());
}
}
void
HAService::startClientAndListener() {
// Add critical section callbacks.
MultiThreadingMgr::instance().addCriticalSectionCallbacks(getCSCallbacksSetName(),
std::bind(&HAService::checkPermissionsClientAndListener, this),
std::bind(&HAService::pauseClientAndListener, this),
std::bind(&HAService::resumeClientAndListener, this));
if (client_) {
client_->start();
}
if (listener_) {
listener_->start();
}
}
void
HAService::pauseClientAndListener() {
// Since this function is used as CS callback all exceptions must be
// suppressed, unlikely though they may be.
try {
if (client_) {
client_->pause();
}
if (listener_) {
listener_->pause();
}
} catch (const std::exception& ex) {
LOG_ERROR(ha_logger, HA_PAUSE_CLIENT_LISTENER_FAILED)
.arg(ex.what());
}
}
void
HAService::resumeClientAndListener() {
// Since this function is used as CS callback all exceptions must be
// suppressed, unlikely though they may be.
try {
if (client_) {
client_->resume();
}
if (listener_) {
listener_->resume();
}
} catch (std::exception& ex) {
LOG_ERROR(ha_logger, HA_RESUME_CLIENT_LISTENER_FAILED)
.arg(config_->getThisServerName())
.arg(ex.what());
}
}
void
HAService::stopClientAndListener() {
// Remove critical section callbacks.
MultiThreadingMgr::instance().removeCriticalSectionCallbacks(getCSCallbacksSetName());
if (client_) {
client_->stop();
}
if (listener_) {
listener_->stop();
}
}
// Explicit instantiations.
template int HAService::getPendingRequest(const Pkt4Ptr&);
template int HAService::getPendingRequest(const Pkt6Ptr&);
} // end of namespace isc::ha
} // end of namespace isc
|