summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorKotresh HR <khiremat@redhat.com>2016-08-01 16:59:33 +0530
committerAravinda VK <avishwan@redhat.com>2016-08-09 00:27:12 -0700
commit949472d7561d3bfd67d8204e433a25dbc8a596cc (patch)
treed73d7439a4214c1fc7bbf01743e59a6aeaa6af18
parentf5fbf46d82e4ea5325e89459c5ad9f753f1ca66b (diff)
features/libgfchangelog: Log failure in gf_histroy_changelog
Add error logs if gf_history_changelog fails. If requested changelog range is not available, log the error and exit instead of continuing the loop and exiting in readdir without logging. Also fixed the duplicate MSGID number in 'changelog-lib-messages.h' Change-Id: Icd71b89ae23b48a71380657ba5649029c32fabfd BUG: 1362151 Signed-off-by: Kotresh HR <khiremat@redhat.com> Reviewed-on: http://review.gluster.org/15064 Smoke: Gluster Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Aravinda VK <avishwan@redhat.com>
-rw-r--r--geo-replication/syncdaemon/libgfchangelog.py5
-rw-r--r--geo-replication/syncdaemon/resource.py5
-rw-r--r--geo-replication/syncdaemon/syncdutils.py2
-rw-r--r--xlators/features/changelog/lib/src/changelog-lib-messages.h11
-rw-r--r--xlators/features/changelog/lib/src/gf-history-changelog.c17
5 files changed, 34 insertions, 6 deletions
diff --git a/geo-replication/syncdaemon/libgfchangelog.py b/geo-replication/syncdaemon/libgfchangelog.py
index 1d580caf6e8..d87b56cd941 100644
--- a/geo-replication/syncdaemon/libgfchangelog.py
+++ b/geo-replication/syncdaemon/libgfchangelog.py
@@ -11,7 +11,7 @@
import os
from ctypes import CDLL, RTLD_GLOBAL, create_string_buffer, get_errno, byref, c_ulong
from ctypes.util import find_library
-from syncdutils import ChangelogException
+from syncdutils import ChangelogException, ChangelogHistoryNotAvailable
class Changes(object):
@@ -99,6 +99,9 @@ class Changes(object):
if ret == -1:
cls.raise_changelog_err()
+ if ret == -2:
+ raise ChangelogHistoryNotAvailable()
+
return (ret, actual_end.value)
@classmethod
diff --git a/geo-replication/syncdaemon/resource.py b/geo-replication/syncdaemon/resource.py
index bc7b13c95bc..ed0e7efe2b2 100644
--- a/geo-replication/syncdaemon/resource.py
+++ b/geo-replication/syncdaemon/resource.py
@@ -36,7 +36,7 @@ import syncdutils
from syncdutils import GsyncdError, select, privileged, boolify, funcode
from syncdutils import umask, entry2pb, gauxpfx, errno_wrap, lstat
from syncdutils import NoPurgeTimeAvailable, PartialHistoryAvailable
-from syncdutils import ChangelogException
+from syncdutils import ChangelogException, ChangelogHistoryNotAvailable
from syncdutils import CHANGELOG_AGENT_CLIENT_VERSION
from gsyncdstatus import GeorepStatus
@@ -1503,6 +1503,9 @@ class GLUSTER(AbstractUrl, SlaveLocal, SlaveRemote):
logging.info('Partial history available, using xsync crawl'
' after consuming history till %s' % str(e))
g1.crawlwrap(oneshot=True, register_time=register_time)
+ except ChangelogHistoryNotAvailable:
+ logging.info('Changelog history not available, using xsync')
+ g1.crawlwrap(oneshot=True, register_time=register_time)
except NoPurgeTimeAvailable:
logging.info('No stime available, using xsync crawl')
g1.crawlwrap(oneshot=True, register_time=register_time)
diff --git a/geo-replication/syncdaemon/syncdutils.py b/geo-replication/syncdaemon/syncdutils.py
index 9d89a3cd415..40eff050a9e 100644
--- a/geo-replication/syncdaemon/syncdutils.py
+++ b/geo-replication/syncdaemon/syncdutils.py
@@ -501,6 +501,8 @@ class NoPurgeTimeAvailable(Exception):
class PartialHistoryAvailable(Exception):
pass
+class ChangelogHistoryNotAvailable(Exception):
+ pass
class ChangelogException(OSError):
pass
diff --git a/xlators/features/changelog/lib/src/changelog-lib-messages.h b/xlators/features/changelog/lib/src/changelog-lib-messages.h
index 976c67f61a9..a6209af71ee 100644
--- a/xlators/features/changelog/lib/src/changelog-lib-messages.h
+++ b/xlators/features/changelog/lib/src/changelog-lib-messages.h
@@ -45,7 +45,7 @@
*/
#define GLFS_COMP_BASE_CHANGELOG_LIB GLFS_MSGID_COMP_CHANGELOG_LIB
-#define GLFS_NUM_MESSAGES 28
+#define GLFS_NUM_MESSAGES 32
#define GLFS_MSGID_END (GLFS_COMP_BASE_CHANGELOG_LIB + GLFS_NUM_MESSAGES + 1)
#define glfs_msg_start_x GLFS_COMP_BASE_CHANGELOG_LIB,\
@@ -280,8 +280,15 @@
* @recommended action
*/
#define CHANGELOG_LIB_MSG_PTHREAD_JOIN_FAILED \
- (GLFS_COMP_BASE_CHANGELOG_LIB + 20)
+ (GLFS_COMP_BASE_CHANGELOG_LIB + 31)
+/*!
+ * @messageid
+ * @diagnosis
+ * @recommended action
+*/
+#define CHANGELOG_LIB_MSG_HIST_FAILED \
+ (GLFS_COMP_BASE_CHANGELOG_LIB + 32)
#define glfs_msg_end_x GLFS_MSGID_END, "Invalid: End of messages"
#endif /* !_CHANGELOG_MESSAGES_H_ */
diff --git a/xlators/features/changelog/lib/src/gf-history-changelog.c b/xlators/features/changelog/lib/src/gf-history-changelog.c
index 5ed50390a7c..45f2d8ba9f3 100644
--- a/xlators/features/changelog/lib/src/gf-history-changelog.c
+++ b/xlators/features/changelog/lib/src/gf-history-changelog.c
@@ -805,6 +805,9 @@ gf_history_changelog (char* changelog_dir, unsigned long start,
ret = pthread_attr_init (&attr);
if (ret != 0) {
+ gf_msg (this->name, GF_LOG_ERROR, errno,
+ CHANGELOG_LIB_MSG_PTHREAD_ERROR,
+ "Pthread init failed");
return -1;
}
@@ -828,6 +831,10 @@ gf_history_changelog (char* changelog_dir, unsigned long start,
/* basic sanity check */
if (start > end || n_parallel <= 0) {
+ gf_msg (this->name, GF_LOG_ERROR, errno,
+ CHANGELOG_LIB_MSG_HIST_FAILED, "Sanity check failed. "
+ "START - %lu END - %lu THREAD_COUNT - %d",
+ start, end, n_parallel);
ret = -1;
goto out;
}
@@ -964,8 +971,14 @@ gf_history_changelog (char* changelog_dir, unsigned long start,
goto out;
- } /* end of range check */
-
+ } else {/* end of range check */
+ gf_msg (this->name, GF_LOG_ERROR, errno,
+ CHANGELOG_LIB_MSG_HIST_FAILED, "Requested changelog "
+ "range is not available. START - %lu CHLOG_MIN - %lu "
+ "CHLOG_MAX - %lu", start, min_ts, max_ts);
+ ret = -2;
+ goto out;
+ }
} /* end of readdir() */
if (!from || !to)
' href='#n433'>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
/*
   Copyright (c) 2011-2012 Red Hat, Inc. <http://www.redhat.com>
   This file is part of GlusterFS.

   This file is licensed to you under your choice of the GNU Lesser
   General Public License, version 3 or any later version (LGPLv3 or
   later), or the GNU General Public License, version 2 (GPLv2), in all
   cases as published by the Free Software Foundation.
*/
#include "common-utils.h"
#include "cli1-xdr.h"
#include "xdr-generic.h"
#include "glusterd.h"
#include "glusterd-op-sm.h"
#include "glusterd-geo-rep.h"
#include "glusterd-store.h"
#include "glusterd-mgmt.h"
#include "glusterd-utils.h"
#include "glusterd-volgen.h"
#include "glusterd-svc-helper.h"
#include "glusterd-messages.h"
#include "glusterd-server-quorum.h"
#include "run.h"
#include "glusterd-volgen.h"
#include <sys/signal.h>

/* misc */

gf_boolean_t
glusterd_is_tiering_supported (char *op_errstr)
{
        xlator_t           *this        = NULL;
        glusterd_conf_t    *conf        = NULL;
        gf_boolean_t        supported   = _gf_false;

        this = THIS;
        GF_VALIDATE_OR_GOTO ("glusterd", this, out);

        conf = this->private;
        GF_VALIDATE_OR_GOTO (this->name, conf, out);

        if (conf->op_version < GD_OP_VERSION_3_7_0)
                goto out;

        supported = _gf_true;

out:
        if (!supported && op_errstr != NULL && conf)
                sprintf (op_errstr, "Tier operation failed. The cluster is "
                         "operating at version %d. Tiering"
                         " is unavailable in this version.",
                         conf->op_version);

        return supported;
}

/* In this function, we decide, based on the 'count' of the brick,
   where to add it in the current volume. 'count' tells us already
   how many of the given bricks are added. other argument are self-
   descriptive. */
int
add_brick_at_right_order (glusterd_brickinfo_t *brickinfo,
                          glusterd_volinfo_t *volinfo, int count,
                          int32_t stripe_cnt, int32_t replica_cnt)
{
        int                   idx     = 0;
        int                   i       = 0;
        int                   sub_cnt = 0;
        glusterd_brickinfo_t *brick   = NULL;

        /* The complexity of the function is in deciding at which index
           to add new brick. Even though it can be defined with a complex
           single formula for all volume, it is separated out to make it
           more readable */
        if (stripe_cnt) {
                /* common formula when 'stripe_count' is set */
                /* idx = ((count / ((stripe_cnt * volinfo->replica_count) -
                   volinfo->dist_leaf_count)) * volinfo->dist_leaf_count) +
                   (count + volinfo->dist_leaf_count);
                */

                sub_cnt = volinfo->dist_leaf_count;

                idx = ((count / ((stripe_cnt * volinfo->replica_count) -
                                 sub_cnt)) * sub_cnt) +
                        (count + sub_cnt);

                goto insert_brick;
        }

        /* replica count is set */
        /* common formula when 'replica_count' is set */
        /* idx = ((count / (replica_cnt - existing_replica_count)) *
           existing_replica_count) +
           (count + existing_replica_count);
        */

        sub_cnt = volinfo->replica_count;
        idx = (count / (replica_cnt - sub_cnt) * sub_cnt) +
                (count + sub_cnt);

insert_brick:
        i = 0;
        cds_list_for_each_entry (brick, &volinfo->bricks, brick_list) {
                i++;
                if (i < idx)
                        continue;
                gf_msg_debug (THIS->name, 0, "brick:%s index=%d, count=%d",
                        brick->path, idx, count);

                cds_list_add (&brickinfo->brick_list, &brick->brick_list);
                break;
        }

        return 0;
}


static int
gd_addbr_validate_stripe_count (glusterd_volinfo_t *volinfo, int stripe_count,
                                int total_bricks, int *type, char *err_str,
                                size_t err_len)
{
        int ret = -1;

        switch (volinfo->type) {
        case GF_CLUSTER_TYPE_NONE:
                if ((volinfo->brick_count * stripe_count) == total_bricks) {
                        /* Change the volume type */
                        *type = GF_CLUSTER_TYPE_STRIPE;
                        gf_msg (THIS->name, GF_LOG_INFO, 0,
                                GD_MSG_VOL_TYPE_CHANGING_INFO,
                                "Changing the type of volume %s from "
                                "'distribute' to 'stripe'", volinfo->volname);
                        ret = 0;
                        goto out;
                } else {
                        snprintf (err_str, err_len, "Incorrect number of "
                                  "bricks (%d) supplied for stripe count (%d).",
                                  (total_bricks - volinfo->brick_count),
                                  stripe_count);
                        gf_msg (THIS->name, GF_LOG_ERROR, EINVAL,
                                GD_MSG_INVALID_ENTRY, "%s", err_str);
                        goto out;
                }
                break;
        case GF_CLUSTER_TYPE_REPLICATE:
                if (!(total_bricks % (volinfo->replica_count * stripe_count))) {
                        /* Change the volume type */
                        *type = GF_CLUSTER_TYPE_STRIPE_REPLICATE;
                        gf_msg (THIS->name, GF_LOG_INFO, 0,
                                GD_MSG_VOL_TYPE_CHANGING_INFO,
                                "Changing the type of volume %s from "
                                "'replicate' to 'replicate-stripe'",
                                volinfo->volname);
                        ret = 0;
                        goto out;
                } else {
                        snprintf (err_str, err_len, "Incorrect number of "
                                  "bricks (%d) supplied for changing volume's "
                                  "stripe count to %d, need at least %d bricks",
                                  (total_bricks - volinfo->brick_count),
                                  stripe_count,
                                  (volinfo->replica_count * stripe_count));
                        gf_msg (THIS->name, GF_LOG_ERROR, EINVAL,
                                GD_MSG_INVALID_ENTRY, "%s", err_str);
                        goto out;
                }
                break;
        case GF_CLUSTER_TYPE_STRIPE:
        case GF_CLUSTER_TYPE_STRIPE_REPLICATE:
                if (stripe_count < volinfo->stripe_count) {
                        snprintf (err_str, err_len,
                                  "Incorrect stripe count (%d) supplied. "
                                  "Volume already has stripe count (%d)",
                                  stripe_count, volinfo->stripe_count);
                        gf_msg (THIS->name, GF_LOG_ERROR, EINVAL,
                                GD_MSG_INVALID_ENTRY, "%s", err_str);
                        goto out;
                }
                if (stripe_count == volinfo->stripe_count) {
                        if (!(total_bricks % volinfo->dist_leaf_count)) {
                                /* its same as the one which exists */
                                ret = 1;
                                goto out;
                        }
                }
                if (stripe_count > volinfo->stripe_count) {
                        /* We have to make sure before and after 'add-brick',
                           the number or subvolumes for distribute will remain
                           same, when stripe count is given */
                        if ((volinfo->brick_count * (stripe_count *
                                                     volinfo->replica_count)) ==
                            (total_bricks * volinfo->dist_leaf_count)) {
                                /* Change the dist_leaf_count */
                                gf_msg (THIS->name, GF_LOG_INFO, 0,
                                        GD_MSG_STRIPE_COUNT_CHANGE_INFO,
                                        "Changing the stripe count of "
                                        "volume %s from %d to %d",
                                        volinfo->volname,
                                        volinfo->stripe_count, stripe_count);
                                ret = 0;
                                goto out;
                        }
                }
                break;
        case GF_CLUSTER_TYPE_DISPERSE:
                snprintf (err_str, err_len, "Volume %s cannot be converted "
                                            "from dispersed to striped-"
                                            "dispersed", volinfo->volname);
                gf_msg(THIS->name, GF_LOG_ERROR, EPERM,
                       GD_MSG_OP_NOT_PERMITTED, "%s", err_str);
                goto out;
        }

out:
        return ret;
}

static int
gd_addbr_validate_replica_count (glusterd_volinfo_t *volinfo, int replica_count,
                                 int arbiter_count, int total_bricks, int *type,
                                 char *err_str, int err_len)
{
        int ret = -1;

        /* replica count is set */
        switch (volinfo->type) {
        case GF_CLUSTER_TYPE_NONE:
                if ((volinfo->brick_count * replica_count) == total_bricks) {
                        /* Change the volume type */
                        *type = GF_CLUSTER_TYPE_REPLICATE;
                        gf_msg (THIS->name, GF_LOG_INFO, 0,
                                GD_MSG_VOL_TYPE_CHANGING_INFO,
                                "Changing the type of volume %s from "
                                "'distribute' to 'replica'", volinfo->volname);
                        ret = 0;
                        goto out;

                } else {
                        snprintf (err_str, err_len, "Incorrect number of "
                                  "bricks (%d) supplied for replica count (%d).",
                                  (total_bricks - volinfo->brick_count),
                                  replica_count);
                        gf_msg (THIS->name, GF_LOG_ERROR, EINVAL,
                                GD_MSG_INVALID_ENTRY, "%s", err_str);
                        goto out;
                }
                break;
        case GF_CLUSTER_TYPE_STRIPE:
                if (!(total_bricks % (volinfo->dist_leaf_count * replica_count))) {
                        /* Change the volume type */
                        *type = GF_CLUSTER_TYPE_STRIPE_REPLICATE;
                        gf_msg (THIS->name, GF_LOG_INFO, 0,
                                GD_MSG_VOL_TYPE_CHANGING_INFO,
                                "Changing the type of volume %s from "
                                "'stripe' to 'replicate-stripe'",
                                volinfo->volname);
                        ret = 0;
                        goto out;
                } else {
                        snprintf (err_str, err_len, "Incorrect number of "
                                  "bricks (%d) supplied for changing volume's "
                                  "replica count to %d, need at least %d "
                                  "bricks",
                                  (total_bricks - volinfo->brick_count),
                                  replica_count, (volinfo->dist_leaf_count *
                                                  replica_count));
                        gf_msg (THIS->name, GF_LOG_ERROR, EINVAL,
                                GD_MSG_INVALID_ENTRY, "%s", err_str);
                        goto out;
                }
                break;
        case GF_CLUSTER_TYPE_REPLICATE:
        case GF_CLUSTER_TYPE_STRIPE_REPLICATE:
                if (replica_count < volinfo->replica_count) {
                        snprintf (err_str, err_len,
                                  "Incorrect replica count (%d) supplied. "
                                  "Volume already has (%d)",
                                  replica_count, volinfo->replica_count);
                        gf_msg (THIS->name, GF_LOG_ERROR, EINVAL,
                                GD_MSG_INVALID_ENTRY, "%s", err_str);
                        goto out;
                }
                if (replica_count == volinfo->replica_count) {
                        if (arbiter_count && !volinfo->arbiter_count) {
                                snprintf (err_str, err_len,
                                          "Cannot convert replica 3 volume "
                                          "to arbiter volume.");
                                gf_msg (THIS->name, GF_LOG_ERROR, EINVAL,
                                        GD_MSG_INVALID_ENTRY, "%s", err_str);
                                goto out;
                        }
                        if (!(total_bricks % volinfo->dist_leaf_count)) {
                                ret = 1;
                                goto out;
                        }
                }
                if (replica_count > volinfo->replica_count) {
                        /* We have to make sure before and after 'add-brick',
                           the number or subvolumes for distribute will remain
                           same, when replica count is given */
                        if ((total_bricks * volinfo->dist_leaf_count) ==
                            (volinfo->brick_count * (replica_count *
                                                     volinfo->stripe_count))) {
                                /* Change the dist_leaf_count */
                                gf_msg (THIS->name, GF_LOG_INFO, 0,
                                        GD_MSG_REPLICA_COUNT_CHANGE_INFO,
                                        "Changing the replica count of "
                                        "volume %s from %d to %d",
                                        volinfo->volname, volinfo->replica_count,
                                        replica_count);
                                ret = 0;
                                goto out;
                        }
                }
                break;
        case GF_CLUSTER_TYPE_DISPERSE:
                snprintf (err_str, err_len, "Volume %s cannot be converted "
                                            "from dispersed to replicated-"
                                            "dispersed", volinfo->volname);
                gf_msg(THIS->name, GF_LOG_ERROR, EPERM,
                       GD_MSG_OP_NOT_PERMITTED, "%s", err_str);
                goto out;
        }
out:
        return ret;
}

static int
gd_rmbr_validate_replica_count (glusterd_volinfo_t *volinfo,
                                int32_t replica_count,
                                int32_t brick_count, char *err_str,
                                size_t err_len)
{
        int ret = -1;
        int replica_nodes = 0;

        switch (volinfo->type) {
        case GF_CLUSTER_TYPE_TIER:
                ret = 1;
                goto out;

        case GF_CLUSTER_TYPE_NONE:
        case GF_CLUSTER_TYPE_STRIPE:
        case GF_CLUSTER_TYPE_DISPERSE:
                snprintf (err_str, err_len,
                          "replica count (%d) option given for non replicate "
                          "volume %s", replica_count, volinfo->volname);
                gf_msg (THIS->name, GF_LOG_WARNING, 0,
                        GD_MSG_VOL_NOT_REPLICA, "%s", err_str);
                goto out;

        case GF_CLUSTER_TYPE_REPLICATE:
        case GF_CLUSTER_TYPE_STRIPE_REPLICATE:
                /* in remove brick, you can only reduce the replica count */
                if (replica_count > volinfo->replica_count) {
                        snprintf (err_str, err_len,
                                  "given replica count (%d) option is more "
                                  "than volume %s's replica count (%d)",
                                  replica_count, volinfo->volname,
                                  volinfo->replica_count);
                        gf_msg (THIS->name, GF_LOG_WARNING, EINVAL,
                                GD_MSG_INVALID_ENTRY, "%s", err_str);
                        goto out;
                }
                if (replica_count == volinfo->replica_count) {
                        /* This means the 'replica N' option on CLI was
                           redundant. Check if the total number of bricks given
                           for removal is same as 'dist_leaf_count' */
                        if (brick_count % volinfo->dist_leaf_count) {
                                snprintf (err_str, err_len,
                                          "number of bricks provided (%d) is "
                                          "not valid. need at least %d "
                                          "(or %dxN)", brick_count,
                                          volinfo->dist_leaf_count,
                                          volinfo->dist_leaf_count);
                                gf_msg (THIS->name, GF_LOG_WARNING, EINVAL,
                                        GD_MSG_INVALID_ENTRY, "%s",
                                        err_str);
                                goto out;
                        }
                        ret = 1;
                        goto out;
                }

                replica_nodes = ((volinfo->brick_count /
                                  volinfo->replica_count) *
                                 (volinfo->replica_count - replica_count));

                if (brick_count % replica_nodes) {
                        snprintf (err_str, err_len,
                                  "need %d(xN) bricks for reducing replica "
                                  "count of the volume from %d to %d",
                                  replica_nodes, volinfo->replica_count,
                                  replica_count);
                        goto out;
                }
                break;
        }

        ret = 0;
out:
        return ret;
}

/* Handler functions */
int
__glusterd_handle_add_brick (rpcsvc_request_t *req)
{
        int32_t                         ret = -1;
        gf_cli_req                      cli_req = {{0,}};
        dict_t                          *dict = NULL;
        char                            *bricks = NULL;
        char                            *volname = NULL;
        int                             brick_count = 0;
        void                            *cli_rsp = NULL;
        char                            err_str[2048] = {0,};
        gf_cli_rsp                      rsp = {0,};
        glusterd_volinfo_t              *volinfo = NULL;
        xlator_t                        *this = NULL;
        int                             total_bricks = 0;
        int32_t                         replica_count = 0;
        int32_t                         arbiter_count = 0;
        int32_t                         stripe_count = 0;
        int                             type = 0;
        glusterd_conf_t                 *conf = NULL;

        this = THIS;
        GF_ASSERT(this);

        GF_ASSERT (req);

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

        ret = xdr_to_generic (req->msg[0], &cli_req,
                              (xdrproc_t)xdr_gf_cli_req);
        if (ret < 0) {
                //failed to decode msg;
                req->rpc_err = GARBAGE_ARGS;
                snprintf (err_str, sizeof (err_str), "Garbage args received");
                goto out;
        }

        gf_msg (this->name, GF_LOG_INFO, 0,
                GD_MSG_ADD_BRICK_REQ_RECVD, "Received add brick req");

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

                ret = dict_unserialize (cli_req.dict.dict_val,
                                        cli_req.dict.dict_len,
                                        &dict);
                if (ret < 0) {
                        gf_msg (this->name, GF_LOG_ERROR, errno,
                                GD_MSG_DICT_UNSERIALIZE_FAIL,
                                "failed to "
                                "unserialize req-buffer to dictionary");
                        snprintf (err_str, sizeof (err_str), "Unable to decode "
                                  "the command");
                        goto out;
                }
        }

        ret = dict_get_str (dict, "volname", &volname);

        if (ret) {
                snprintf (err_str, sizeof (err_str), "Unable to get volume "
                          "name");
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "%s", err_str);
                goto out;
        }

        if (!(ret = glusterd_check_volume_exists (volname))) {
                ret = -1;
                snprintf (err_str, sizeof (err_str), "Volume %s does not exist",
                          volname);
                gf_msg (this->name, GF_LOG_ERROR, EINVAL,
                        GD_MSG_VOL_NOT_FOUND, "%s", err_str);
                goto out;
        }

        ret = dict_get_int32 (dict, "count", &brick_count);
        if (ret) {
                snprintf (err_str, sizeof (err_str), "Unable to get volume "
                          "brick count");
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "%s", err_str);
                goto out;
        }

        ret = dict_get_int32 (dict, "replica-count", &replica_count);
        if (!ret) {
                gf_msg (this->name, GF_LOG_INFO, errno,
                        GD_MSG_DICT_GET_SUCCESS, "replica-count is %d",
                        replica_count);
        }

        ret = dict_get_int32 (dict, "arbiter-count", &arbiter_count);
        if (!ret) {
                gf_msg (this->name, GF_LOG_INFO, errno,
                        GD_MSG_DICT_GET_SUCCESS, "arbiter-count is %d",
                        arbiter_count);
        }

        ret = dict_get_int32 (dict, "stripe-count", &stripe_count);
        if (!ret) {
                gf_msg (this->name, GF_LOG_INFO, errno,
                        GD_MSG_DICT_GET_SUCCESS, "stripe-count is %d",
                        stripe_count);
        }

        if (!dict_get (dict, "force")) {
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "Failed to get flag");
                goto out;
        }

        ret = glusterd_volinfo_find (volname, &volinfo);
        if (ret) {
                snprintf (err_str, sizeof (err_str), "Unable to get volinfo "
                          "for volume name %s", volname);
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_VOLINFO_GET_FAIL, "%s", err_str);
                goto out;

        }

        total_bricks = volinfo->brick_count + brick_count;

        if (dict_get (dict, "attach-tier")) {
                if (volinfo->type == GF_CLUSTER_TYPE_TIER) {
                        snprintf (err_str, sizeof (err_str),
                                  "Volume %s is already a tier.", volname);
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_VOL_ALREADY_TIER, "%s", err_str);
                        ret = -1;
                        goto out;
                }

                if (glusterd_is_tiering_supported(err_str) == _gf_false) {
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_VERSION_UNSUPPORTED,
                                "Tiering not supported at this version");
                        ret = -1;
                        goto out;
                }

                ret = dict_get_int32 (dict, "hot-type", &type);
                if (ret) {
                        gf_msg (this->name, GF_LOG_ERROR, errno,
                                GD_MSG_DICT_GET_FAILED,
                                "failed to get type from dictionary");
                        goto out;
                }

                goto brick_val;
        }

        ret = glusterd_disallow_op_for_tier (volinfo, GD_OP_ADD_BRICK, -1);
        if (ret) {
                snprintf (err_str, sizeof (err_str), "Add-brick operation is "
                          "not supported on a tiered volume %s", volname);
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_OP_UNSUPPORTED, "%s", err_str);
                goto out;
        }

        if (!stripe_count && !replica_count) {
                if (volinfo->type == GF_CLUSTER_TYPE_NONE)
                        goto brick_val;

                if ((volinfo->brick_count < volinfo->dist_leaf_count) &&
                    (total_bricks <= volinfo->dist_leaf_count))
                        goto brick_val;

                if ((brick_count % volinfo->dist_leaf_count) != 0) {
                        snprintf (err_str, sizeof (err_str), "Incorrect number "
                                  "of bricks supplied %d with count %d",
                                 brick_count, volinfo->dist_leaf_count);
                        gf_msg (this->name, GF_LOG_ERROR, EINVAL,
                                GD_MSG_VOL_NOT_REPLICA, "%s", err_str);
                        ret = -1;
                        goto out;
                }
                goto brick_val;
                /* done with validation.. below section is if stripe|replica
                   count is given */
        }

        /* These bricks needs to be added one per a replica or stripe volume */
        if (stripe_count) {
                ret = gd_addbr_validate_stripe_count (volinfo, stripe_count,
                                                      total_bricks, &type,
                                                      err_str,
                                                      sizeof (err_str));
                if (ret == -1) {
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_COUNT_VALIDATE_FAILED, "%s", err_str);
                        goto out;
                }

                /* if stripe count is same as earlier, set it back to 0 */
                if (ret == 1)
                        stripe_count = 0;

                ret = dict_set_int32 (dict, "stripe-count", stripe_count);
                if (ret) {
                        gf_msg (this->name, GF_LOG_ERROR, errno,
                                GD_MSG_DICT_GET_FAILED,
                                "failed to set the stripe-count in dict");
                        goto out;
                }
                goto brick_val;
        }

        ret = gd_addbr_validate_replica_count (volinfo, replica_count,
                                               arbiter_count, total_bricks,
                                               &type, err_str,
                                               sizeof (err_str));
        if (ret == -1) {
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_COUNT_VALIDATE_FAILED, "%s", err_str);
                goto out;
        }

        /* if replica count is same as earlier, set it back to 0 */
        if (ret == 1)
                replica_count = 0;

        ret = dict_set_int32 (dict, "replica-count", replica_count);
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_SET_FAILED,
                        "failed to set the replica-count in dict");
                goto out;
        }

brick_val:
        ret = dict_get_str (dict, "bricks", &bricks);
        if (ret) {
                snprintf (err_str, sizeof (err_str), "Unable to get volume "
                          "bricks");
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "%s", err_str);
                goto out;
        }

        if (type != volinfo->type) {
                ret = dict_set_int32 (dict, "type", type);
                if (ret) {
                        gf_msg (this->name, GF_LOG_ERROR, errno,
                                GD_MSG_DICT_SET_FAILED,
                                "failed to set the new type in dict");
                        goto out;
                }
        }

        if (conf->op_version <= GD_OP_VERSION_3_7_5) {
                gf_msg_debug (this->name, 0, "The cluster is operating at "
                          "version less than or equal to %d. Falling back "
                          "to syncop framework.",
                          GD_OP_VERSION_3_7_5);
                ret = glusterd_op_begin_synctask (req, GD_OP_ADD_BRICK, dict);
        } else {
                ret = glusterd_mgmt_v3_initiate_all_phases (req,
                                                            GD_OP_ADD_BRICK,
                                                            dict);
        }

out:
        if (ret) {
                rsp.op_ret = -1;
                rsp.op_errno = 0;
                if (err_str[0] == '\0')
                        snprintf (err_str, sizeof (err_str), "Operation failed");
                rsp.op_errstr = err_str;
                cli_rsp = &rsp;
                glusterd_to_cli (req, cli_rsp, NULL, 0, NULL,
                                 (xdrproc_t)xdr_gf_cli_rsp, dict);
                ret = 0; //sent error to cli, prevent second reply
        }

        free (cli_req.dict.dict_val); //its malloced by xdr

        return ret;
}

int
glusterd_handle_add_brick (rpcsvc_request_t *req)
{
        return glusterd_big_locked_handler (req, __glusterd_handle_add_brick);
}

static int
subvol_matcher_init (int **subvols, int count)
{
        int ret = -1;

        *subvols = GF_CALLOC (count, sizeof(int), gf_gld_mt_int);
        if (*subvols)
                ret = 0;

        return ret;
}

static void
subvol_matcher_update (int *subvols, glusterd_volinfo_t *volinfo,
                       glusterd_brickinfo_t *brickinfo)
{
        glusterd_brickinfo_t *tmp        = NULL;
        int32_t               sub_volume = 0;
        int                   pos        = 0;

        cds_list_for_each_entry (tmp, &volinfo->bricks, brick_list) {

                if (strcmp (tmp->hostname, brickinfo->hostname) ||
                    strcmp (tmp->path, brickinfo->path)) {
                        pos++;
                        continue;
                }
                gf_msg_debug (THIS->name, 0, LOGSTR_FOUND_BRICK,
                                        brickinfo->hostname, brickinfo->path,
                                        volinfo->volname);
                sub_volume = (pos / volinfo->dist_leaf_count);
                subvols[sub_volume]++;
                break;
        }

}

static int
subvol_matcher_verify (int *subvols, glusterd_volinfo_t *volinfo, char *err_str,
                       size_t err_len, char *vol_type, int replica_count)
{
        int i = 0;
        int ret = 0;
        int count = volinfo->replica_count-replica_count;

        if (replica_count) {
                for (i = 0; i < volinfo->subvol_count; i++) {
                        if (subvols[i] != count) {
                                ret = -1;
                                snprintf (err_str, err_len, "Remove exactly %d"
                                " brick(s) from each subvolume.", count);
                                break;
                        }
                }
                return ret;
        }

       do {

                if (subvols[i] % volinfo->dist_leaf_count == 0) {
                        continue;
                } else {
                        ret = -1;
                        snprintf (err_str, err_len,
                                "Bricks not from same subvol for %s", vol_type);
                        break;
                }
        } while (++i < volinfo->subvol_count);

        return ret;
}

static void
subvol_matcher_destroy (int *subvols)
{
        GF_FREE (subvols);
}

static int
glusterd_set_detach_bricks(dict_t *dict, glusterd_volinfo_t *volinfo)
{
        char key[256] = {0,};
        char value[256] = {0,};
        int brick_num = 0;
        int hot_brick_num = 0;
        glusterd_brickinfo_t *brickinfo;
        int ret = 0;

        /* cold tier bricks at tail of list so use reverse iteration */
        cds_list_for_each_entry_reverse (brickinfo, &volinfo->bricks,
                                         brick_list) {
                brick_num++;
                if (brick_num > volinfo->tier_info.cold_brick_count) {
                        hot_brick_num++;
                        sprintf (key, "brick%d", hot_brick_num);
                        snprintf (value, 256, "%s:%s",
                                  brickinfo->hostname,
                                  brickinfo->path);

                        ret = dict_set_str (dict, key, strdup(value));
                        if (ret)
                                break;
                }
        }

        ret = dict_set_int32(dict, "count", hot_brick_num);
        if (ret)
                return -1;

        return hot_brick_num;
}

static int
glusterd_remove_brick_validate_arbiters (glusterd_volinfo_t *volinfo,
                                         int32_t count, int32_t replica_count,
                                         glusterd_brickinfo_t **brickinfo_list,
                                         char *err_str, size_t err_len)
{
        int i = 0;
        int ret = 0;
        glusterd_brickinfo_t *brickinfo = NULL;
        glusterd_brickinfo_t *last = NULL;
        char *arbiter_array = NULL;

        if ((volinfo->type != GF_CLUSTER_TYPE_REPLICATE) &&
            (volinfo->type != GF_CLUSTER_TYPE_STRIPE_REPLICATE))
                goto out;

        if (!replica_count || !volinfo->arbiter_count)
                goto out;

        if (replica_count == 2) {
                /* If it is an arbiter to replica 2 conversion, only permit
                *  removal of the arbiter brick.*/
                for (i = 0; i < count; i++) {
                        brickinfo = brickinfo_list[i];
                        last = get_last_brick_of_brick_group (volinfo,
                                                              brickinfo);
                        if (last != brickinfo) {
                                snprintf (err_str, err_len, "Remove arbiter "
                                          "brick(s) only when converting from "
                                           "arbiter to replica 2 subvolume.");
                                ret = -1;
                                goto out;
                        }
                }
        } else if (replica_count == 1) {
                /* If it is an arbiter to plain distribute conversion, in every
                 * replica subvol, the arbiter has to be one of the bricks that
                 * are removed. */
                arbiter_array = GF_CALLOC (volinfo->subvol_count,
                                           sizeof (*arbiter_array),
                                           gf_common_mt_char);
                if (!arbiter_array)
                        return -1;
                for (i = 0; i < count; i++) {
                        brickinfo = brickinfo_list[i];
                        last = get_last_brick_of_brick_group (volinfo,
                                                              brickinfo);
                        if (last == brickinfo)
                                arbiter_array[brickinfo->group] = 1;
                }
                for (i = 0; i < volinfo->subvol_count; i++)
                        if (!arbiter_array[i]) {
                                snprintf (err_str, err_len, "Removed bricks "
                                          "must contain arbiter when converting"
                                           " to plain distrubute.");
                                ret = -1;
                                break;
                        }
                GF_FREE (arbiter_array);
        }

out:
        return ret;
}

int
__glusterd_handle_remove_brick (rpcsvc_request_t *req)
{
        int32_t                   ret              = -1;
        gf_cli_req                cli_req          = {{0,}};
        dict_t                   *dict             = NULL;
        int32_t                   count            = 0;
        char                     *brick            = NULL;
        char                      key[256]         = {0,};
        int                       i                = 1;
        glusterd_volinfo_t       *volinfo          = NULL;
        glusterd_brickinfo_t     *brickinfo        = NULL;
        glusterd_brickinfo_t     **brickinfo_list  = NULL;
        int                      *subvols          = NULL;
        char                      err_str[2048]    = {0};
        gf_cli_rsp                rsp              = {0,};
        void                     *cli_rsp          = NULL;
        char                      vol_type[256]    = {0,};
        int32_t                   replica_count    = 0;
        char                     *volname          = 0;
        xlator_t                 *this             = NULL;
        int                       cmd              = -1;

        GF_ASSERT (req);
        this = THIS;
        GF_ASSERT (this);

        ret = xdr_to_generic (req->msg[0], &cli_req,
                              (xdrproc_t)xdr_gf_cli_req);
        if (ret < 0) {
                //failed to decode msg;
                req->rpc_err = GARBAGE_ARGS;
                snprintf (err_str, sizeof (err_str), "Received garbage args");
                goto out;
        }

        gf_msg (this->name, GF_LOG_INFO, 0,
                GD_MSG_REM_BRICK_REQ_RECVD,
                "Received rem brick req");

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

                ret = dict_unserialize (cli_req.dict.dict_val,
                                        cli_req.dict.dict_len,
                                        &dict);
                if (ret < 0) {
                        gf_msg (this->name, GF_LOG_ERROR, errno,
                                GD_MSG_DICT_UNSERIALIZE_FAIL,
                                "failed to "
                                "unserialize req-buffer to dictionary");
                        snprintf (err_str, sizeof (err_str), "Unable to decode "
                                  "the command");
                        goto out;
                }
        }

        ret = dict_get_str (dict, "volname", &volname);
        if (ret) {
                snprintf (err_str, sizeof (err_str), "Unable to get volume "
                          "name");
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "%s", err_str);
                goto out;
        }

        ret = dict_get_int32 (dict, "count", &count);
        if (ret) {
                snprintf (err_str, sizeof (err_str), "Unable to get brick "
                          "count");
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "%s", err_str);
                goto out;
        }

        ret = glusterd_volinfo_find (volname, &volinfo);
        if (ret) {
                 snprintf (err_str, sizeof (err_str),"Volume %s does not exist",
                           volname);
                 gf_msg (this->name, GF_LOG_ERROR, EINVAL,
                         GD_MSG_VOL_NOT_FOUND, "%s", err_str);
                 goto out;
        }

        if ((volinfo->type == GF_CLUSTER_TYPE_TIER) &&
            (glusterd_is_tiering_supported(err_str) == _gf_false)) {
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_VERSION_UNSUPPORTED,
                        "Tiering not supported at this version");
                ret = -1;
                goto out;
        }

        ret = dict_get_int32 (dict, "command", &cmd);
        if (ret) {
                snprintf (err_str, sizeof (err_str), "Unable to get cmd "
                          "ccommand");
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "%s", err_str);
                goto out;
        }

        ret = glusterd_disallow_op_for_tier (volinfo, GD_OP_REMOVE_BRICK, cmd);
        if (ret) {
                snprintf (err_str, sizeof (err_str),
                          "Removing brick from a Tier volume is not allowed");
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_OP_UNSUPPORTED, "%s", err_str);
                goto out;
        }

        ret = dict_get_int32 (dict, "replica-count", &replica_count);
        if (!ret) {
                gf_msg (this->name, GF_LOG_INFO, errno,
                        GD_MSG_DICT_GET_FAILED,
                        "request to change replica-count to %d", replica_count);
                ret = gd_rmbr_validate_replica_count (volinfo, replica_count,
                                                      count, err_str,
                                                      sizeof (err_str));
                if (ret < 0) {
                        /* logging and error msg are done in above function
                           itself */
                        goto out;
                }
                dict_del (dict, "replica-count");
                if (ret) {
                        replica_count = 0;
                } else {
                        ret = dict_set_int32 (dict, "replica-count",
                                              replica_count);
                        if (ret) {
                                gf_msg (this->name, GF_LOG_WARNING, errno,
                                        GD_MSG_DICT_SET_FAILED,
                                        "failed to set the replica_count "
                                        "in dict");
                                goto out;
                        }
                }
        }

        /* 'vol_type' is used for giving the meaning full error msg for user */
        if (volinfo->type == GF_CLUSTER_TYPE_REPLICATE) {
                strcpy (vol_type, "replica");
        } else if (volinfo->type == GF_CLUSTER_TYPE_STRIPE) {
                strcpy (vol_type, "stripe");
        } else if (volinfo->type == GF_CLUSTER_TYPE_STRIPE_REPLICATE) {
                strcpy (vol_type, "stripe-replicate");
        } else if (volinfo->type == GF_CLUSTER_TYPE_DISPERSE) {
                strcpy (vol_type, "disperse");
        } else {
                strcpy (vol_type, "distribute");
        }

	/* Do not allow remove-brick if the volume is a stripe volume*/
	if ((volinfo->type == GF_CLUSTER_TYPE_STRIPE) &&
            (volinfo->brick_count == volinfo->stripe_count)) {
                snprintf (err_str, sizeof (err_str),
                          "Removing brick from a stripe volume is not allowed");
                gf_msg (this->name, GF_LOG_ERROR, EPERM,
                        GD_MSG_OP_NOT_PERMITTED, "%s", err_str);
                ret = -1;
                goto out;
	}

	if (!replica_count &&
            (volinfo->type == GF_CLUSTER_TYPE_STRIPE_REPLICATE) &&
            (volinfo->brick_count == volinfo->dist_leaf_count)) {
                snprintf (err_str, sizeof(err_str),
                          "Removing bricks from stripe-replicate"
                          " configuration is not allowed without reducing "
                          "replica or stripe count explicitly.");
                gf_msg (this->name, GF_LOG_ERROR, EPERM,
                        GD_MSG_OP_NOT_PERMITTED_AC_REQD, "%s", err_str);
                ret = -1;
                goto out;
        }

	if (!replica_count &&
            (volinfo->type == GF_CLUSTER_TYPE_REPLICATE) &&
            (volinfo->brick_count == volinfo->dist_leaf_count)) {
                snprintf (err_str, sizeof (err_str),
                          "Removing bricks from replicate configuration "
                          "is not allowed without reducing replica count "
                          "explicitly.");
                gf_msg (this->name, GF_LOG_ERROR, EPERM,
                        GD_MSG_OP_NOT_PERMITTED_AC_REQD, "%s", err_str);
                ret = -1;
                goto out;
        }

	/* Do not allow remove-brick if the bricks given is less than
           the replica count or stripe count */
        if (!replica_count && (volinfo->type != GF_CLUSTER_TYPE_NONE) &&
            (volinfo->type != GF_CLUSTER_TYPE_TIER))  {
                if (volinfo->dist_leaf_count &&
                    (count % volinfo->dist_leaf_count)) {
                        snprintf (err_str, sizeof (err_str), "Remove brick "
                                  "incorrect brick count of %d for %s %d",
                                  count, vol_type, volinfo->dist_leaf_count);
                        gf_msg (this->name, GF_LOG_ERROR, EINVAL,
                                GD_MSG_INVALID_ENTRY, "%s", err_str);
                        ret = -1;
                        goto out;
                }
        }

        /* subvol match is not required for tiered volume*/
        if ((volinfo->type != GF_CLUSTER_TYPE_NONE) &&
             (volinfo->type != GF_CLUSTER_TYPE_TIER) &&
             (volinfo->subvol_count > 1)) {
                ret = subvol_matcher_init (&subvols, volinfo->subvol_count);
                if (ret)
                        goto out;
        }

        if (volinfo->type == GF_CLUSTER_TYPE_TIER)
                count = glusterd_set_detach_bricks(dict, volinfo);

        brickinfo_list = GF_CALLOC (count, sizeof (*brickinfo_list),
                                    gf_common_mt_pointer);
        if (!brickinfo_list) {
                ret = -1;
                goto out;
        }

        while ( i <= count) {
                snprintf (key, sizeof (key), "brick%d", i);
                ret = dict_get_str (dict, key, &brick);
                if (ret) {
                        snprintf (err_str, sizeof (err_str), "Unable to get %s",
                                  key);
                        gf_msg (this->name, GF_LOG_ERROR, errno,
                                GD_MSG_DICT_GET_FAILED, "%s", err_str);
                        goto out;
                }
                gf_msg_debug (this->name, 0, "Remove brick count %d brick:"
                        " %s", i, brick);

                ret = glusterd_volume_brickinfo_get_by_brick(brick, volinfo,
                                                             &brickinfo,
                                                             _gf_false);

                if (ret) {
                        snprintf (err_str, sizeof (err_str), "Incorrect brick "
                                  "%s for volume %s", brick, volname);
                        gf_msg (this->name, GF_LOG_ERROR, EINVAL,
                                GD_MSG_BRICK_NOT_FOUND, "%s", err_str);
                        goto out;
                }
                brickinfo_list[i-1] = brickinfo;

                i++;
                if ((volinfo->type == GF_CLUSTER_TYPE_NONE) ||
                    (volinfo->brick_count <= volinfo->dist_leaf_count))
                        continue;

                /* Find which subvolume the brick belongs to.
                 * subvol match is not required for tiered volume
                 *
                 */
                if (volinfo->type != GF_CLUSTER_TYPE_TIER)
                        subvol_matcher_update (subvols, volinfo, brickinfo);
        }

        /* Check if the bricks belong to the same subvolumes.*/
        /* subvol match is not required for tiered volume*/
         if ((volinfo->type != GF_CLUSTER_TYPE_NONE) &&
             (volinfo->type != GF_CLUSTER_TYPE_TIER) &&
             (volinfo->subvol_count > 1)) {
                ret = subvol_matcher_verify (subvols, volinfo,
                                             err_str, sizeof(err_str),
                                             vol_type, replica_count);
                if (ret)
                        goto out;
        }

        ret = glusterd_remove_brick_validate_arbiters (volinfo, count,
                                                       replica_count,
                                                       brickinfo_list,
                                                       err_str,
                                                       sizeof (err_str));
        if (ret)
                goto out;

        ret = glusterd_op_begin_synctask (req, GD_OP_REMOVE_BRICK, dict);

out:
        if (ret) {
                rsp.op_ret = -1;
                rsp.op_errno = 0;
                if (err_str[0] == '\0')
                        snprintf (err_str, sizeof (err_str),
                                  "Operation failed");
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_GLUSTERD_OP_FAILED, "%s", err_str);
                rsp.op_errstr = err_str;
                cli_rsp = &rsp;
                glusterd_to_cli (req, cli_rsp, NULL, 0, NULL,
                                 (xdrproc_t)xdr_gf_cli_rsp, dict);

                ret = 0; //sent error to cli, prevent second reply

        }

        if (brickinfo_list)
                GF_FREE (brickinfo_list);
        subvol_matcher_destroy (subvols);
        free (cli_req.dict.dict_val); //its malloced by xdr

        return ret;
}

int
glusterd_handle_remove_brick (rpcsvc_request_t *req)
{
        return glusterd_big_locked_handler (req,
                                            __glusterd_handle_remove_brick);
}

static int
_glusterd_restart_gsync_session (dict_t *this, char *key,
                                 data_t *value, void *data)
{
        char                          *slave      = NULL;
        char                          *slave_buf  = NULL;
        char                          *path_list  = NULL;
        char                          *slave_vol  = NULL;
        char                          *slave_host = NULL;
        char                          *slave_url  = NULL;
        char                          *conf_path  = NULL;
        char                         **errmsg     = NULL;
        int                            ret        = -1;
        glusterd_gsync_status_temp_t  *param      = NULL;
        gf_boolean_t                   is_running = _gf_false;

        param = (glusterd_gsync_status_temp_t *)data;

        GF_ASSERT (param);
        GF_ASSERT (param->volinfo);

        slave = strchr(value->data, ':');
        if (slave) {
                slave++;
                slave_buf = gf_strdup (slave);
                if (!slave_buf) {
                        gf_msg ("glusterd", GF_LOG_ERROR, ENOMEM,
                                GD_MSG_NO_MEMORY,
                                "Failed to gf_strdup");
                        ret = -1;
                        goto out;
                }
        }
        else
                return 0;

        ret = dict_set_dynstr (param->rsp_dict, "slave", slave_buf);
        if (ret) {
                gf_msg ("glusterd", GF_LOG_ERROR, errno,
                        GD_MSG_DICT_SET_FAILED,
                        "Unable to store slave");
                if (slave_buf)
                        GF_FREE(slave_buf);
                goto out;
        }

        ret = glusterd_get_slave_details_confpath (param->volinfo,
                                                   param->rsp_dict, &slave_url,
                                                   &slave_host, &slave_vol,
                                                   &conf_path, errmsg);
        if (ret) {
                if (*errmsg)
                        gf_msg ("glusterd", GF_LOG_ERROR, 0,
                                GD_MSG_SLAVE_CONFPATH_DETAILS_FETCH_FAIL,
                                "%s", *errmsg);
                else
                        gf_msg ("glusterd", GF_LOG_ERROR, 0,
                                GD_MSG_SLAVE_CONFPATH_DETAILS_FETCH_FAIL,
                                "Unable to fetch slave or confpath details.");
                goto out;
        }

        /* In cases that gsyncd is not running, we will not invoke it
         * because of add-brick. */
        ret = glusterd_check_gsync_running_local (param->volinfo->volname,
                                                  slave, conf_path,
                                                  &is_running);
        if (ret) {
                gf_msg ("glusterd", GF_LOG_ERROR, 0,
                        GD_MSG_GSYNC_VALIDATION_FAIL, "gsync running validation failed.");
                goto out;
        }
        if (_gf_false == is_running) {
                gf_msg_debug ("glusterd", 0, "gsync session for %s and %s is"
                        " not running on this node. Hence not restarting.",
                        param->volinfo->volname, slave);
                ret = 0;
                goto out;
        }

        ret = glusterd_get_local_brickpaths (param->volinfo, &path_list);
        if (!path_list) {
                gf_msg_debug ("glusterd", 0, "This node not being part of"
                        " volume should not be running gsyncd. Hence"
                        " no gsyncd process to restart.");
                ret = 0;
                goto out;
        }

        ret = glusterd_check_restart_gsync_session (param->volinfo, slave,
                                                    param->rsp_dict, path_list,
                                                    conf_path, 0);
        if (ret)
                gf_msg ("glusterd", GF_LOG_ERROR, 0,
                        GD_MSG_GSYNC_RESTART_FAIL,
                        "Unable to restart gsync session.");

out:
        gf_msg_debug ("glusterd", 0, "Returning %d.", ret);
        return ret;
}

/* op-sm */

int
glusterd_op_perform_add_bricks (glusterd_volinfo_t *volinfo, int32_t count,
                                char  *bricks, dict_t *dict)
{
        char                         *brick          = NULL;
        int32_t                       i              = 1;
        char                         *brick_list     = NULL;
        char                         *free_ptr1      = NULL;
        char                         *free_ptr2      = NULL;
        char                         *saveptr        = NULL;
        int32_t                       ret            = -1;
        int32_t                       stripe_count   = 0;
        int32_t                       replica_count  = 0;
        int32_t                       arbiter_count  = 0;
        int32_t                       type           = 0;
        glusterd_brickinfo_t         *brickinfo      = NULL;
        glusterd_gsync_status_temp_t  param          = {0, };
        gf_boolean_t                  restart_needed = 0;
        char                          msg[1024] __attribute__((unused)) = {0, };
        int                           caps           = 0;
        int                           brickid        = 0;
        char                          key[PATH_MAX]  = "";
        char                         *brick_mount_dir  = NULL;
        xlator_t                     *this           = NULL;
        glusterd_conf_t              *conf           = NULL;
        gf_boolean_t                  is_valid_add_brick = _gf_false;

        this = THIS;
        GF_ASSERT (this);
        GF_ASSERT (volinfo);

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

        if (bricks) {
                brick_list = gf_strdup (bricks);
                free_ptr1 = brick_list;
        }

        if (count)
                brick = strtok_r (brick_list+1, " \n", &saveptr);

        if (dict) {
                ret = dict_get_int32 (dict, "stripe-count", &stripe_count);
                if (!ret)
                        gf_msg (THIS->name, GF_LOG_INFO, errno,
                                GD_MSG_DICT_GET_SUCCESS,
                                "stripe-count is set %d", stripe_count);

                ret = dict_get_int32 (dict, "replica-count", &replica_count);
                if (!ret)
                        gf_msg (THIS->name, GF_LOG_INFO, errno,
                                GD_MSG_DICT_GET_SUCCESS,
                                "replica-count is set %d", replica_count);
                ret = dict_get_int32 (dict, "arbiter-count", &arbiter_count);
                if (!ret)
                        gf_msg (THIS->name, GF_LOG_INFO, errno,
                                GD_MSG_DICT_GET_SUCCESS,
                                "arbiter-count is set %d", arbiter_count);
                ret = dict_get_int32 (dict, "type", &type);
                if (!ret)
                        gf_msg (THIS->name, GF_LOG_INFO, errno,
                                GD_MSG_DICT_GET_SUCCESS,
                                "type is set %d, need to change it", type);
        }

        brickid = glusterd_get_next_available_brickid (volinfo);
        if (brickid < 0)
                goto out;
        while ( i <= count) {
                ret = glusterd_brickinfo_new_from_brick (brick, &brickinfo,
                                                         _gf_true, NULL);
                if (ret)
                        goto out;

                GLUSTERD_ASSIGN_BRICKID_TO_BRICKINFO (brickinfo, volinfo,
                                                      brickid++);

                /* A bricks mount dir is required only by snapshots which were
                 * introduced in gluster-3.6.0
                 */
                if (conf->op_version >= GD_OP_VERSION_3_6_0) {
                        brick_mount_dir = NULL;

                        snprintf (key, sizeof(key), "brick%d.mount_dir", i);
                        ret = dict_get_str (dict, key, &brick_mount_dir);
                        if (ret) {
                                gf_msg (this->name, GF_LOG_ERROR, errno,
                                        GD_MSG_DICT_GET_FAILED,
                                        "%s not present", key);
                                goto out;
                        }
                        strncpy (brickinfo->mount_dir, brick_mount_dir,
                                 sizeof(brickinfo->mount_dir));
                }

                ret = glusterd_resolve_brick (brickinfo);
                if (ret)
                        goto out;

                /* hot tier bricks are added to head of brick list */
                if (dict_get (dict, "attach-tier")) {
                        cds_list_add (&brickinfo->brick_list, &volinfo->bricks);
                } else if (stripe_count || replica_count) {
                        add_brick_at_right_order (brickinfo, volinfo, (i - 1),
                                                  stripe_count, replica_count);
                } else {
                        cds_list_add_tail (&brickinfo->brick_list,
                                           &volinfo->bricks);
                }
                brick = strtok_r (NULL, " \n", &saveptr);
                i++;
                volinfo->brick_count++;

        }

        /* Gets changed only if the options are given in add-brick cli */
        if (type)
                volinfo->type = type;

        if (replica_count) {
                volinfo->replica_count = replica_count;
        }
        if (arbiter_count) {
                volinfo->arbiter_count = arbiter_count;
        }
        if (stripe_count) {
                volinfo->stripe_count = stripe_count;
        }
        volinfo->dist_leaf_count = glusterd_get_dist_leaf_count (volinfo);

        /* backward compatibility */
        volinfo->sub_count = ((volinfo->dist_leaf_count == 1) ? 0:
                              volinfo->dist_leaf_count);

        volinfo->subvol_count = (volinfo->brick_count /
                                 volinfo->dist_leaf_count);

        ret = 0;
        if (GLUSTERD_STATUS_STARTED != volinfo->status)
                goto generate_volfiles;

        ret = generate_brick_volfiles (volinfo);
        if (ret)
                goto out;

        brick_list = gf_strdup (bricks);
        free_ptr2 = brick_list;
        i = 1;

        if (count)
                brick = strtok_r (brick_list+1, " \n", &saveptr);
#ifdef HAVE_BD_XLATOR
        if (brickinfo->vg[0])
                caps = CAPS_BD | CAPS_THIN |
                        CAPS_OFFLOAD_COPY | CAPS_OFFLOAD_SNAPSHOT;
#endif

        /* This check needs to be added to distinguish between
         * attach-tier commands and add-brick commands.
         * When a tier is attached, adding is done via add-brick
         * and setting of pending xattrs shouldn't be done for
         * attach-tiers as they are virtually new volumes.
         */
        if (glusterd_is_volume_replicate (volinfo)) {
                if (replica_count &&
                    !dict_get (dict, "attach-tier") &&
                    conf->op_version >= GD_OP_VERSION_3_7_10) {
                        is_valid_add_brick = _gf_true;
                        ret = generate_dummy_client_volfiles (volinfo);
                        if (ret) {
                                gf_msg (THIS->name, GF_LOG_ERROR, 0,
                                        GD_MSG_VOLFILE_CREATE_FAIL,
                                        "Failed to create volfile.");
                                goto out;
                                }
                        }
        }

        while (i <= count) {
                ret = glusterd_volume_brickinfo_get_by_brick (brick, volinfo,
                                                              &brickinfo,
                                                             _gf_true);
                if (ret)
                        goto out;
#ifdef HAVE_BD_XLATOR
                /* Check for VG/thin pool if its BD volume */
                if (brickinfo->vg[0]) {
                        ret = glusterd_is_valid_vg (brickinfo, 0, msg);
                        if (ret) {
                                gf_msg (THIS->name, GF_LOG_CRITICAL, 0,
                                        GD_MSG_INVALID_VG, "%s", msg);
                                goto out;
                        }
                        /* if anyone of the brick does not have thin support,
                           disable it for entire volume */
                        caps &= brickinfo->caps;
                } else
                        caps = 0;
#endif

                if (gf_uuid_is_null (brickinfo->uuid)) {
                        ret = glusterd_resolve_brick (brickinfo);
                        if (ret) {
                                gf_msg ("glusterd", GF_LOG_ERROR, 0,
                                        GD_MSG_RESOLVE_BRICK_FAIL, FMTSTR_RESOLVE_BRICK,
                                        brickinfo->hostname, brickinfo->path);
                                goto out;
                        }
                }

                /* if the volume is a replicate volume, do: */
                if (is_valid_add_brick) {
                        if (!gf_uuid_compare (brickinfo->uuid, MY_UUID)) {
                                ret = glusterd_handle_replicate_brick_ops (
                                                           volinfo, brickinfo,
                                                           GD_OP_ADD_BRICK);
                                if (ret < 0)
                                        goto out;
                        }
                }
                ret = glusterd_brick_start (volinfo, brickinfo,
                                            _gf_true);
                if (ret)
                        goto out;
                i++;
                brick = strtok_r (NULL, " \n", &saveptr);

                /* Check if the brick is added in this node, and set
                 * the restart_needed flag. */
                if ((!gf_uuid_compare (brickinfo->uuid, MY_UUID)) &&
                    !restart_needed) {
                        restart_needed = 1;
                        gf_msg_debug ("glusterd", 0,
                                "Restart gsyncd session, if it's already "
                                "running.");
                }
        }

        /* If the restart_needed flag is set, restart gsyncd sessions for that
         * particular master with all the slaves. */
        if (restart_needed) {
                param.rsp_dict = dict;
                param.volinfo = volinfo;
                dict_foreach (volinfo->gsync_slaves,
                              _glusterd_restart_gsync_session, &param);
        }
        volinfo->caps = caps;

generate_volfiles:
        if (conf->op_version <= GD_OP_VERSION_3_7_5) {
               ret = glusterd_create_volfiles_and_notify_services (volinfo);
        } else {
                /*
                 * The cluster is operating at version greater than
                 * gluster-3.7.5. So no need to sent volfile fetch
                 * request in commit phase, the same will be done
                 * in post validate phase with v3 framework.
                 */
        }

out:
        GF_FREE (free_ptr1);
        GF_FREE (free_ptr2);

        gf_msg_debug ("glusterd", 0, "Returning %d", ret);
        return ret;
}


int
glusterd_op_perform_remove_brick (glusterd_volinfo_t  *volinfo, char *brick,
                                  int force, int *need_migrate)
{
        glusterd_brickinfo_t    *brickinfo = NULL;
        int32_t                  ret = -1;
        glusterd_conf_t         *priv = NULL;

        GF_ASSERT (volinfo);
        GF_ASSERT (brick);

        priv = THIS->private;
        GF_ASSERT (priv);

        ret = glusterd_volume_brickinfo_get_by_brick (brick, volinfo,
                                                      &brickinfo,
                                                      _gf_false);
        if (ret)
                goto out;

        ret = glusterd_resolve_brick (brickinfo);
        if (ret)
                goto out;

        glusterd_volinfo_reset_defrag_stats (volinfo);

        if (!gf_uuid_compare (brickinfo->uuid, MY_UUID)) {
                /* Only if the brick is in this glusterd, do the rebalance */
                if (need_migrate)
                        *need_migrate = 1;
        }

        if (force) {
                ret = glusterd_brick_stop (volinfo, brickinfo,
                                           _gf_true);
                if (ret) {
                        gf_msg (THIS->name, GF_LOG_ERROR, 0,
                                GD_MSG_BRICK_STOP_FAIL, "Unable to stop "
                                "glusterfs, ret: %d", ret);
                }
                goto out;
        }

        brickinfo->decommissioned = 1;
        ret = 0;
out:
        gf_msg_debug ("glusterd", 0, "Returning %d", ret);
        return ret;
}

int
glusterd_op_stage_add_brick (dict_t *dict, char **op_errstr, dict_t *rsp_dict)
{
        int                                     ret = 0;
        char                                    *volname = NULL;
        int                                     count = 0;
        int                                     replica_count = 0;
        int                                     arbiter_count = 0;
        int                                     i = 0;
        int32_t                                 local_brick_count = 0;
        char                                    *bricks    = NULL;
        char                                    *brick_list = NULL;
        char                                    *saveptr = NULL;
        char                                    *free_ptr = NULL;
        char                                    *brick = NULL;
        glusterd_brickinfo_t                    *brickinfo = NULL;
        glusterd_volinfo_t                      *volinfo = NULL;
        xlator_t                                *this = NULL;
        char                                    msg[2048] = {0,};
        char                                    key[PATH_MAX] = "";
        gf_boolean_t                            brick_alloc = _gf_false;
        char                                    *all_bricks = NULL;
        char                                    *str_ret = NULL;
        gf_boolean_t                            is_force = _gf_false;
        glusterd_conf_t                         *conf = NULL;

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

        ret = dict_get_str (dict, "volname", &volname);
        if (ret) {
                gf_msg (THIS->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED,
                        "Unable to get volume name");
                goto out;
        }

        ret = glusterd_volinfo_find (volname, &volinfo);
        if (ret) {
                gf_msg (THIS->name, GF_LOG_ERROR, 0,
                        GD_MSG_VOL_NOT_FOUND,
                        "Unable to find volume: %s", volname);
                goto out;
        }

        ret = glusterd_validate_volume_id (dict, volinfo);
        if (ret)
                goto out;

        ret = dict_get_int32 (dict, "replica-count", &replica_count);
        if (ret) {
                gf_msg_debug (THIS->name, 0,
                        "Unable to get replica count");
        }

        ret = dict_get_int32 (dict, "arbiter-count", &arbiter_count);
        if (ret) {
                gf_msg_debug (THIS->name, 0,
                        "No arbiter count present in the dict");
        }

        if (replica_count > 0) {
                ret = op_version_check (this, GD_OP_VER_PERSISTENT_AFR_XATTRS,
                                        msg, sizeof(msg));
                if (ret) {
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_OP_VERSION_MISMATCH, "%s", msg);
                        *op_errstr = gf_strdup (msg);
                        goto out;
                }
        }

        if (glusterd_is_volume_replicate (volinfo)) {
                /* Do not allow add-brick for stopped volumes when replica-count
                 * is being increased.
                 */
                if (conf->op_version >= GD_OP_VERSION_3_7_10 &&
                    !dict_get (dict, "attach-tier") &&
                    replica_count &&
                    GLUSTERD_STATUS_STOPPED == volinfo->status) {
                        ret = -1;
                        snprintf (msg, sizeof (msg), " Volume must not be in"
                                  " stopped state when replica-count needs to "
                                  " be increased.");
                        gf_msg (THIS->name, GF_LOG_ERROR, 0,
                                GD_MSG_BRICK_ADD_FAIL, "%s", msg);
                        *op_errstr = gf_strdup (msg);
                        goto out;
                }
                /* op-version check for replica 2 to arbiter conversion. If we
                * dont have this check, an older peer added as arbiter brick
                * will not have the  arbiter xlator in its volfile. */
                if ((conf->op_version < GD_OP_VERSION_3_8_0) &&
                    (arbiter_count == 1) && (replica_count == 3)) {
                        ret = -1;
                        snprintf (msg, sizeof (msg), "Cluster op-version must "
                                  "be >= 30800 to add arbiter brick to a "
                                  "replica 2 volume.");
                        gf_msg (THIS->name, GF_LOG_ERROR, 0,
                                GD_MSG_BRICK_ADD_FAIL, "%s", msg);
                        *op_errstr = gf_strdup (msg);
                        goto out;
                }
        }

        if (conf->op_version > GD_OP_VERSION_3_7_5 &&
            is_origin_glusterd (dict)) {
                ret = glusterd_validate_quorum (this, GD_OP_ADD_BRICK, dict,
                                                op_errstr);
                if (ret) {
                        gf_msg (this->name, GF_LOG_CRITICAL, 0,
                                GD_MSG_SERVER_QUORUM_NOT_MET,
                                "Server quorum not met. Rejecting operation.");
                        goto out;
                }
        } else {
                /* Case 1: conf->op_version <= GD_OP_VERSION_3_7_5
                 *         in this case the add-brick is running
                 *         syncop framework that will do a quorum
                 *         check by default
                 * Case 2: We don't need to do quorum check on every
                 *         node, only originator glusterd need to
                 *         check for quorum
                 * So nothing need to be done in else
                 */
        }

        if (glusterd_is_defrag_on(volinfo)) {
                snprintf (msg, sizeof(msg), "Volume name %s rebalance is in "
                          "progress. Please retry after completion", volname);
                gf_msg (THIS->name, GF_LOG_ERROR, 0,
                        GD_MSG_OIP_RETRY_LATER, "%s", msg);
                *op_errstr = gf_strdup (msg);
                ret = -1;
                goto out;
        }

        if (dict_get(dict, "attach-tier")) {

                /*
                 * This check is needed because of add/remove brick
                 * is not supported on a tiered volume. So once a tier
                 * is attached we cannot commit or stop the remove-brick
                 * task. Please change this comment once we start supporting
                 * add/remove brick on a tiered volume.
                 */
                if (!gd_is_remove_brick_committed (volinfo)) {

                        snprintf (msg, sizeof (msg), "An earlier remove-brick "
                                  "task exists for volume %s. Either commit it"
                                  " or stop it before attaching a tier.",
                                  volinfo->volname);
                        gf_msg (THIS->name, GF_LOG_ERROR, 0,
                                GD_MSG_OLD_REMOVE_BRICK_EXISTS, "%s", msg);
                        *op_errstr = gf_strdup (msg);
                        ret = -1;
                        goto out;
                }
        }

        ret = dict_get_int32 (dict, "count", &count);
        if (ret) {
                gf_msg ("glusterd", GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "Unable to get count");
                goto out;
        }

        ret = dict_get_str (dict, "bricks", &bricks);
        if (ret) {
                gf_msg (THIS->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "Unable to get bricks");
                goto out;
        }

        is_force = dict_get_str_boolean (dict, "force", _gf_false);

        if (bricks) {
                brick_list = gf_strdup (bricks);
                all_bricks = gf_strdup (bricks);
                free_ptr = brick_list;
        }

        if (count)
                brick = strtok_r (brick_list+1, " \n", &saveptr);


        while ( i < count) {
                if (!glusterd_store_is_valid_brickpath (volname, brick) ||
                    !glusterd_is_valid_volfpath (volname, brick)) {
                        snprintf (msg, sizeof (msg), "brick path %s is "
                                  "too long", brick);
                        gf_msg (THIS->name, GF_LOG_ERROR, 0,
                                GD_MSG_BRKPATH_TOO_LONG, "%s", msg);
                        *op_errstr = gf_strdup (msg);

                        ret = -1;
                        goto out;

                }

                ret = glusterd_brickinfo_new_from_brick (brick, &brickinfo,
                                                         _gf_true, NULL);
                if (ret) {
                        gf_msg (THIS->name, GF_LOG_ERROR, 0,
                                GD_MSG_BRICK_NOT_FOUND,
                                "Add-brick: Unable"
                                " to get brickinfo");
                        goto out;
                }
                brick_alloc = _gf_true;

                ret = glusterd_new_brick_validate (brick, brickinfo, msg,
                                                   sizeof (msg));
                if (ret) {
                        *op_errstr = gf_strdup (msg);
                        ret = -1;
                        goto out;
                }

                if (!gf_uuid_compare (brickinfo->uuid, MY_UUID)) {
#ifdef HAVE_BD_XLATOR
                        if (brickinfo->vg[0]) {
                                ret = glusterd_is_valid_vg (brickinfo, 1, msg);
                                if (ret) {
                                        gf_msg (THIS->name, GF_LOG_ERROR, EINVAL,
                                               GD_MSG_INVALID_VG, "%s",
                                                msg);
                                        *op_errstr = gf_strdup (msg);
                                        goto out;
                                }
                        }
#endif

                        ret = glusterd_validate_and_create_brickpath (brickinfo,
                                                          volinfo->volume_id,
                                                          op_errstr, is_force);
                        if (ret)
                                goto out;

                        /* A bricks mount dir is required only by snapshots which were
                         * introduced in gluster-3.6.0
                         */
                        if (conf->op_version >= GD_OP_VERSION_3_6_0) {
                                ret = glusterd_get_brick_mount_dir
                                        (brickinfo->path, brickinfo->hostname,
                                         brickinfo->mount_dir);
                                if (ret) {
                                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                                GD_MSG_BRICK_MOUNTDIR_GET_FAIL,
                                                "Failed to get brick mount_dir");
                                        goto out;
                                }

                                snprintf (key, sizeof(key), "brick%d.mount_dir",
                                          i + 1);
                                ret = dict_set_dynstr_with_alloc
                                        (rsp_dict, key, brickinfo->mount_dir);
                                if (ret) {
                                        gf_msg (this->name, GF_LOG_ERROR, errno,
                                                GD_MSG_DICT_SET_FAILED,
                                                "Failed to set %s", key);
                                        goto out;
                                }
                        }

                        local_brick_count = i + 1;
                }

                glusterd_brickinfo_delete (brickinfo);
                brick_alloc = _gf_false;
                brickinfo = NULL;
                brick = strtok_r (NULL, " \n", &saveptr);
                i++;
        }

        ret = dict_set_int32 (rsp_dict, "brick_count",
                              local_brick_count);
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_SET_FAILED,
                        "Failed to set local_brick_count");
                goto out;
        }

out:
        GF_FREE (free_ptr);
        if (brick_alloc && brickinfo)
                glusterd_brickinfo_delete (brickinfo);
        GF_FREE (str_ret);
        GF_FREE (all_bricks);

        gf_msg_debug (THIS->name, 0, "Returning %d", ret);

        return ret;
}

static int
glusterd_remove_brick_validate_bricks (gf1_op_commands cmd, int32_t brick_count,
                                       dict_t *dict,
                                       glusterd_volinfo_t *volinfo,
                                       char **errstr)
{
        char                   *brick       = NULL;
        char                    msg[2048]   = {0,};
        char                    key[256]    = {0,};
        glusterd_brickinfo_t   *brickinfo   = NULL;
        glusterd_peerinfo_t    *peerinfo    = NULL;
        int                     i           = 0;
        int                     ret         = -1;

        /* Check whether all the nodes of the bricks to be removed are
        * up, if not fail the operation */
        for (i = 1; i <= brick_count; i++) {
                snprintf (key, sizeof (key), "brick%d", i);
                ret = dict_get_str (dict, key, &brick);
                if (ret) {
                        snprintf (msg, sizeof (msg),
                                  "Unable to get %s", key);
                        *errstr = gf_strdup (msg);
                        goto out;
                }

                ret =
                glusterd_volume_brickinfo_get_by_brick(brick, volinfo,
                                                       &brickinfo,
                                                       _gf_false);
                if (ret) {
                        snprintf (msg, sizeof (msg), "Incorrect brick "
                                  "%s for volume %s", brick, volinfo->volname);
                        *errstr = gf_strdup (msg);
                        goto out;
                }
                /* Do not allow commit if the bricks are not decommissioned
                 * if its a remove brick commit or detach-tier commit
                 */
                if (!brickinfo->decommissioned) {
                        if (cmd == GF_OP_CMD_COMMIT) {
                                snprintf (msg, sizeof (msg), "Brick %s "
                                          "is not decommissioned. "
                                          "Use start or force option", brick);
                                *errstr = gf_strdup (msg);
                                ret = -1;
                                goto out;
                        }

                        if (cmd == GF_OP_CMD_DETACH_COMMIT) {
                                snprintf (msg, sizeof (msg), "Bricks in Hot "
                                          "tier are not decommissioned yet. Use "
                                          "gluster volume tier <VOLNAME> "
                                          "detach start to start the decommission process");
                                *errstr = gf_strdup (msg);
                                ret = -1;
                                goto out;
                        }
                } else {
                        if (cmd == GF_OP_CMD_DETACH_COMMIT &&
                            (volinfo->rebal.defrag_status == GF_DEFRAG_STATUS_STARTED)) {
                                        snprintf (msg, sizeof (msg), "Bricks in Hot "
                                                  "tier are not decommissioned yet. Wait for "
                                                  "the detach to complete using gluster volume "
                                                  "tier <VOLNAME> status.");
                                        *errstr = gf_strdup (msg);
                                        ret = -1;
                                        goto out;
                            }
                }

                if (glusterd_is_local_brick (THIS, volinfo, brickinfo)) {
                        if (((cmd == GF_OP_CMD_START) ||
                            (cmd == GF_OP_CMD_DETACH_START))  &&
                            brickinfo->status != GF_BRICK_STARTED) {
                                snprintf (msg, sizeof (msg), "Found stopped "
                                          "brick %s", brick);
                                *errstr = gf_strdup (msg);
                                ret = -1;
                                goto out;
                        }
                        continue;
                }

                rcu_read_lock ();
                peerinfo = glusterd_peerinfo_find_by_uuid
                                                (brickinfo->uuid);
                if (!peerinfo) {
                        snprintf (msg, sizeof(msg), "Host node of the "
                                  "brick %s is not in cluster", brick);
                        *errstr = gf_strdup (msg);
                        ret = -1;
                        rcu_read_unlock ();
                        goto out;
                }
                if (!peerinfo->connected) {
                        snprintf (msg, sizeof(msg), "Host node of the "
                                  "brick %s is down", brick);
                        *errstr = gf_strdup (msg);
                        ret = -1;
                        rcu_read_unlock ();
                        goto out;
                }
                rcu_read_unlock ();
        }

out:
        return ret;
}

int
glusterd_op_stage_remove_brick (dict_t *dict, char **op_errstr)
{
        int                     ret         = -1;
        char                   *volname     = NULL;
        glusterd_volinfo_t     *volinfo     = NULL;
        char                   *errstr      = NULL;
        int32_t                 brick_count = 0;
        char                    msg[2048]   = {0,};
        int32_t                 flag        = 0;
        gf1_op_commands         cmd         = GF_OP_CMD_NONE;
        char                   *task_id_str = NULL;
        xlator_t               *this        = NULL;
        gsync_status_param_t    param       = {0,};

        this = THIS;
        GF_ASSERT (this);

        ret = op_version_check (this, GD_OP_VER_PERSISTENT_AFR_XATTRS,
                                msg, sizeof(msg));
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_OP_VERSION_MISMATCH, "%s", msg);
                *op_errstr = gf_strdup (msg);
                goto out;
        }

        ret = dict_get_str (dict, "volname", &volname);
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_SET_FAILED, "Unable to get volume name");
                goto out;
        }

        ret = glusterd_volinfo_find (volname, &volinfo);

        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_VOL_NOT_FOUND, "Volume %s does not exist", volname);
                goto out;
        }

        ret = glusterd_validate_volume_id (dict, volinfo);
        if (ret)
                goto out;

        ret = dict_get_int32 (dict, "command", &flag);
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED,
                        "Unable to get brick command");
                goto out;
        }
        cmd = flag;

        ret = dict_get_int32 (dict, "count", &brick_count);
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "Unable to get brick count");
                goto out;
        }

        ret = 0;
        if (volinfo->brick_count == brick_count) {
                errstr = gf_strdup ("Deleting all the bricks of the "
                                    "volume is not allowed");
                ret = -1;
                goto out;
        }

        ret = -1;
        switch (cmd) {
        case GF_OP_CMD_NONE:
                errstr = gf_strdup ("no remove-brick command issued");
                goto out;

        case GF_OP_CMD_STATUS:
                ret = 0;
                goto out;

        case GF_OP_CMD_DETACH_START:
                if (volinfo->type != GF_CLUSTER_TYPE_TIER) {
                        snprintf (msg, sizeof(msg), "volume %s is not a tier "
                                  "volume", volinfo->volname);
                        errstr = gf_strdup (msg);
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_VOL_NOT_TIER, "%s", errstr);
                        goto out;
                }

        case GF_OP_CMD_START:
        {
                if ((volinfo->type == GF_CLUSTER_TYPE_REPLICATE) &&
                    dict_get (dict, "replica-count")) {
                        snprintf (msg, sizeof(msg), "Migration of data is not "
                                  "needed when reducing replica count. Use the"
                                  " 'force' option");
                        errstr = gf_strdup (msg);
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_USE_THE_FORCE, "%s", errstr);
                        goto out;
                }

                if (GLUSTERD_STATUS_STARTED != volinfo->status) {
                        if (volinfo->type == GF_CLUSTER_TYPE_TIER) {
                                snprintf (msg, sizeof (msg), "Volume %s needs "
                                          "to be started before detach-tier "
                                          "(you can use 'force' or 'commit' "
                                          "to override this behavior)",
                                          volinfo->volname);
                        } else {
                                snprintf (msg, sizeof (msg), "Volume %s needs "
                                          "to be started before remove-brick "
                                          "(you can use 'force' or 'commit' "
                                          "to override this behavior)",
                                          volinfo->volname);
                        }
                        errstr = gf_strdup (msg);
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_VOL_NOT_STARTED, "%s", errstr);
                        goto out;
                }
                if (!gd_is_remove_brick_committed (volinfo)) {
                        snprintf (msg, sizeof (msg), "An earlier remove-brick "
                                  "task exists for volume %s. Either commit it"
                                  " or stop it before starting a new task.",
                                  volinfo->volname);
                        errstr = gf_strdup (msg);
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_OLD_REMOVE_BRICK_EXISTS, "Earlier remove-brick"
                                " task exists for volume %s.",
                                volinfo->volname);
                        goto out;
                }
                if (glusterd_is_defrag_on(volinfo)) {
                        errstr = gf_strdup("Rebalance is in progress. Please "
                                           "retry after completion");
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_OIP_RETRY_LATER, "%s", errstr);
                        goto out;
                }

                /* Check if the connected clients are all of version
                 * glusterfs-3.6 and higher. This is needed to prevent some data
                 * loss issues that could occur when older clients are connected
                 * when rebalance is run.
                 */
                ret = glusterd_check_client_op_version_support
                        (volname, GD_OP_VERSION_3_6_0, NULL);
                if (ret) {
                        ret = gf_asprintf (op_errstr, "Volume %s has one or "
                                           "more connected clients of a version"
                                           " lower than GlusterFS-v3.6.0. "
                                           "Starting remove-brick in this state "
                                           "could lead to data loss.\nPlease "
                                           "disconnect those clients before "
                                           "attempting this command again.",
                                           volname);
                        goto out;
                }

                ret = glusterd_remove_brick_validate_bricks (cmd, brick_count,
                                                             dict, volinfo,
                                                             &errstr);
                if (ret)
                        goto out;

                if (is_origin_glusterd (dict)) {
                        ret = glusterd_generate_and_set_task_id
                                (dict, GF_REMOVE_BRICK_TID_KEY);
                        if (ret) {
                                gf_msg (this->name, GF_LOG_ERROR, 0,
                                        GD_MSG_TASKID_GEN_FAIL,
                                        "Failed to generate task-id");
                                goto out;
                        }
                } else {
                        ret = dict_get_str (dict, GF_REMOVE_BRICK_TID_KEY,
                                            &task_id_str);
                        if (ret) {
                                gf_msg (this->name, GF_LOG_WARNING, errno,
                                        GD_MSG_DICT_GET_FAILED,
                                        "Missing remove-brick-id");
                                ret = 0;
                        }
                }
                break;
        }

        case GF_OP_CMD_STOP:
        case GF_OP_CMD_STOP_DETACH_TIER:
                ret = 0;
                break;

        case GF_OP_CMD_DETACH_COMMIT:
                if (volinfo->type != GF_CLUSTER_TYPE_TIER) {
                        snprintf (msg, sizeof(msg), "volume %s is not a tier "
                                  "volume", volinfo->volname);
                        errstr = gf_strdup (msg);
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_VOL_NOT_TIER, "%s", errstr);
                        goto out;
                }
                ret = glusterd_remove_brick_validate_bricks (cmd, brick_count,
                                                             dict, volinfo,
                                                             &errstr);
                if (ret)
                        goto out;

                /* If geo-rep is configured, for this volume, it should be
                 * stopped.
                 */
                param.volinfo = volinfo;
                ret = glusterd_check_geo_rep_running (&param, op_errstr);
                if (ret || param.is_active) {
                        ret = -1;
                        goto out;
                }
                break;

        case GF_OP_CMD_COMMIT:
                if (volinfo->decommission_in_progress) {
                        errstr = gf_strdup ("use 'force' option as migration "
                                            "is in progress");
                        goto out;
                }

                if (volinfo->rebal.defrag_status == GF_DEFRAG_STATUS_FAILED) {
                        errstr = gf_strdup ("use 'force' option as migration "
                                            "has failed");
                        goto out;
                }

                ret = glusterd_remove_brick_validate_bricks (cmd, brick_count,
                                                             dict, volinfo,
                                                             &errstr);
                if (ret)
                        goto out;

                /* If geo-rep is configured, for this volume, it should be
                 * stopped.
                 */
                param.volinfo = volinfo;
                ret = glusterd_check_geo_rep_running (&param, op_errstr);
                if (ret || param.is_active) {
                        ret = -1;
                        goto out;
                }

                break;

        case GF_OP_CMD_DETACH_COMMIT_FORCE:
                if (volinfo->type != GF_CLUSTER_TYPE_TIER) {
                        snprintf (msg, sizeof(msg), "volume %s is not a tier "
                                  "volume", volinfo->volname);
                        errstr = gf_strdup (msg);
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_VOL_NOT_TIER, "%s", errstr);
                        goto out;
                }
        case GF_OP_CMD_COMMIT_FORCE:
                break;
        }
        ret = 0;

out:
        gf_msg_debug (this->name, 0, "Returning %d", ret);
        if (ret && errstr) {
                if (op_errstr)
                        *op_errstr = errstr;
        }

        return ret;
}

int
glusterd_remove_brick_migrate_cbk (glusterd_volinfo_t *volinfo,
                                   gf_defrag_status_t status)
{
        int                   ret = 0;

#if 0  /* TODO: enable this behavior once cluster-wide awareness comes for
          defrag cbk function */
        glusterd_brickinfo_t *brickinfo = NULL;
        glusterd_brickinfo_t *tmp = NULL;

        switch (status) {
        case GF_DEFRAG_STATUS_PAUSED:
        case GF_DEFRAG_STATUS_FAILED:
                /* No changes required in the volume file.
                   everything should remain as is */
                break;
        case GF_DEFRAG_STATUS_STOPPED:
                /* Fall back to the old volume file */
                cds_list_for_each_entry_safe (brickinfo, tmp, &volinfo->bricks,
                                              brick_list) {
                        if (!brickinfo->decommissioned)
                                continue;
                        brickinfo->decommissioned = 0;
                }
                break;

        case GF_DEFRAG_STATUS_COMPLETE:
                /* Done with the task, you can remove the brick from the
                   volume file */
                cds_list_for_each_entry_safe (brickinfo, tmp, &volinfo->bricks,
                                              brick_list) {
                        if (!brickinfo->decommissioned)
                                continue;
                        gf_log (THIS->name, GF_LOG_INFO, "removing the brick %s",
                                brickinfo->path);
                        brickinfo->decommissioned = 0;
                        if (GLUSTERD_STATUS_STARTED == volinfo->status) {
                            /*TODO: use the 'atomic' flavour of brick_stop*/
                                ret = glusterd_brick_stop (volinfo, brickinfo);
                                if (ret) {
                                        gf_log (THIS->name, GF_LOG_ERROR,
                                                "Unable to stop glusterfs (%d)", ret);
                                }
                        }
                        glusterd_delete_brick (volinfo, brickinfo);
                }
                break;

        default:
                GF_ASSERT (!"cbk function called with wrong status");
                break;
        }

        ret = glusterd_create_volfiles_and_notify_services (volinfo);
        if (ret)
                gf_log (THIS->name, GF_LOG_ERROR,
                        "Unable to write volume files (%d)", ret);

        ret = glusterd_store_volinfo (volinfo, GLUSTERD_VOLINFO_VER_AC_INCREMENT);
        if (ret)
                gf_log (THIS->name, GF_LOG_ERROR,
                        "Unable to store volume info (%d)", ret);


        if (GLUSTERD_STATUS_STARTED == volinfo->status) {
                ret = glusterd_check_generate_start_nfs ();
                if (ret)
                        gf_log (THIS->name, GF_LOG_ERROR,
                                "Unable to start nfs process (%d)", ret);
        }

#endif

        volinfo->decommission_in_progress = 0;
        return ret;
}

static int
glusterd_op_perform_attach_tier (dict_t *dict,
                                 glusterd_volinfo_t *volinfo,
                                 int count,
                                 char *bricks)
{
        int                                     ret = 0;
        int                                     replica_count = 0;
        int                                     type = 0;

        /*
         * Store the new (cold) tier's structure until the graph is generated.
         * If there is a failure before the graph is generated the
         * structure will revert to its original state.
         */
        volinfo->tier_info.cold_dist_leaf_count = volinfo->dist_leaf_count;
        volinfo->tier_info.cold_type           = volinfo->type;
        volinfo->tier_info.cold_brick_count    = volinfo->brick_count;
        volinfo->tier_info.cold_replica_count  = volinfo->replica_count;
        volinfo->tier_info.cold_disperse_count = volinfo->disperse_count;
        volinfo->tier_info.cold_redundancy_count = volinfo->redundancy_count;

        ret = dict_get_int32 (dict, "replica-count", &replica_count);
        if (!ret)
                volinfo->tier_info.hot_replica_count  = replica_count;
        else
                volinfo->tier_info.hot_replica_count  = 1;
        volinfo->tier_info.hot_brick_count     = count;
        ret = dict_get_int32 (dict, "hot-type", &type);
        volinfo->tier_info.hot_type      = type;
        ret = dict_set_int32 (dict, "type", GF_CLUSTER_TYPE_TIER);

        if (!ret)
                ret = dict_set_str (volinfo->dict, "features.ctr-enabled", "on");

        if (!ret)
                ret = dict_set_str (volinfo->dict, "cluster.tier-mode", "cache");

        return ret;
}

int
glusterd_op_add_brick (dict_t *dict, char **op_errstr)
{
        int                                     ret = 0;
        char                                    *volname = NULL;
        glusterd_conf_t                         *priv = NULL;
        glusterd_volinfo_t                      *volinfo = NULL;
        xlator_t                                *this = NULL;
        char                                    *bricks = NULL;
        int32_t                                 count = 0;

        this = THIS;
        GF_ASSERT (this);

        priv = this->private;
        GF_ASSERT (priv);

        ret = dict_get_str (dict, "volname", &volname);

        if (ret) {
                gf_msg ("glusterd", GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "Unable to get volume name");
                goto out;
        }

        ret = glusterd_volinfo_find (volname, &volinfo);

        if (ret) {
                gf_msg ("glusterd", GF_LOG_ERROR, EINVAL,
                        GD_MSG_VOL_NOT_FOUND, "Unable to allocate memory");
                goto out;
        }

        ret = dict_get_int32 (dict, "count", &count);
        if (ret) {
                gf_msg ("glusterd", GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "Unable to get count");
                goto out;
        }


        ret = dict_get_str (dict, "bricks", &bricks);
        if (ret) {
                gf_msg ("glusterd", GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "Unable to get bricks");
                goto out;
        }

        if (dict_get(dict, "attach-tier")) {
                gf_msg_debug (THIS->name, 0, "Adding tier");
                glusterd_op_perform_attach_tier (dict, volinfo, count, bricks);
        }

        ret = glusterd_op_perform_add_bricks (volinfo, count, bricks, dict);
        if (ret) {
                gf_msg ("glusterd", GF_LOG_ERROR, 0,
                        GD_MSG_BRICK_ADD_FAIL, "Unable to add bricks");
                goto out;
        }
        if (priv->op_version <= GD_OP_VERSION_3_7_5) {
               ret = glusterd_store_volinfo (volinfo,
                                             GLUSTERD_VOLINFO_VER_AC_INCREMENT);
                if (ret)
                        goto out;
        } else {
                 /*
                 * The cluster is operating at version greater than
                 * gluster-3.7.5. So no need to store volfiles
                 * in commit phase, the same will be done
                 * in post validate phase with v3 framework.
                 */
        }

        if (GLUSTERD_STATUS_STARTED == volinfo->status)
                ret = glusterd_svcs_manager (volinfo);

out:
        return ret;
}

static void
glusterd_op_perform_detach_tier (glusterd_volinfo_t *volinfo)
{
        volinfo->type             = volinfo->tier_info.cold_type;
        volinfo->replica_count    = volinfo->tier_info.cold_replica_count;
        volinfo->disperse_count   = volinfo->tier_info.cold_disperse_count;
        volinfo->redundancy_count = volinfo->tier_info.cold_redundancy_count;
        volinfo->dist_leaf_count  = volinfo->tier_info.cold_dist_leaf_count;
}

int
glusterd_op_remove_brick (dict_t *dict, char **op_errstr)
{
        int                     ret            = -1;
        char                    *volname       = NULL;
        glusterd_volinfo_t      *volinfo       = NULL;
        char                    *brick         = NULL;
        int32_t                 count          = 0;
        int32_t                 i              = 1;
        char                    key[256]       = {0,};
        int32_t                 flag           = 0;
        char                    err_str[4096]  = {0,};
        int                     need_rebalance = 0;
        int                     force          = 0;
        gf1_op_commands         cmd            = 0;
        int32_t                 replica_count  = 0;
        glusterd_brickinfo_t    *brickinfo     = NULL;
        glusterd_brickinfo_t    *tmp           = NULL;
        char                    *task_id_str   = NULL;
        xlator_t                *this          = NULL;
        dict_t                  *bricks_dict   = NULL;
        char                    *brick_tmpstr  = NULL;
        int                      start_remove  = 0;
        uint32_t                 commit_hash   = 0;
        int                      defrag_cmd    = 0;
        int                      detach_commit = 0;
        void                    *tier_info     = NULL;
        char                    *cold_shd_key  = NULL;
        char                    *hot_shd_key   = NULL;
        int                      delete_key    = 1;

        this = THIS;
        GF_ASSERT (this);

        ret = dict_get_str (dict, "volname", &volname);

        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_BRICK_ADD_FAIL, "Unable to get volume name");
                goto out;
        }

        ret = glusterd_volinfo_find (volname, &volinfo);
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, EINVAL,
                        GD_MSG_VOL_NOT_FOUND, "Unable to allocate memory");
                goto out;
        }

        ret = dict_get_int32 (dict, "command", &flag);
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "Unable to get command");
                goto out;
        }
        cmd = flag;

        if ((GF_OP_CMD_START == cmd) ||
            (GF_OP_CMD_DETACH_START == cmd))
                start_remove = 1;

        /* Set task-id, if available, in ctx dict for operations other than
         * start
         */

        if (is_origin_glusterd (dict) && (!start_remove)) {
                if (!gf_uuid_is_null (volinfo->rebal.rebalance_id)) {
                        ret = glusterd_copy_uuid_to_dict
                                (volinfo->rebal.rebalance_id, dict,
                                 GF_REMOVE_BRICK_TID_KEY);
                        if (ret) {
                                gf_msg (this->name, GF_LOG_ERROR, 0,
                                        GD_MSG_REMOVE_BRICK_ID_SET_FAIL,
                                        "Failed to set remove-brick-id");
                                goto out;
                        }
                }
        }

        /* Clear task-id, rebal.op and stored bricks on commmitting/stopping
         * remove-brick */
        if ((!start_remove) && (cmd != GF_OP_CMD_STATUS)) {
                gf_uuid_clear (volinfo->rebal.rebalance_id);
                volinfo->rebal.op = GD_OP_NONE;
                dict_unref (volinfo->rebal.dict);
                volinfo->rebal.dict = NULL;
        }

        ret = -1;
        switch (cmd) {
        case GF_OP_CMD_NONE:
                goto out;

        case GF_OP_CMD_STATUS:
                ret = 0;
                goto out;

        case GF_OP_CMD_STOP:
        case GF_OP_CMD_STOP_DETACH_TIER:
        {
                /* Fall back to the old volume file */
                cds_list_for_each_entry_safe (brickinfo, tmp, &volinfo->bricks,
                                              brick_list) {
                        if (!brickinfo->decommissioned)
                                continue;
                        brickinfo->decommissioned = 0;
                }
                ret = glusterd_create_volfiles_and_notify_services (volinfo);
                if (ret) {
                        gf_msg (this->name, GF_LOG_WARNING, 0,
                                GD_MSG_VOLFILE_CREATE_FAIL,
                                "failed to create volfiles");
                        goto out;
                }

                ret = glusterd_store_volinfo (volinfo,
                                             GLUSTERD_VOLINFO_VER_AC_INCREMENT);
                if (ret) {
                        gf_msg (this->name, GF_LOG_WARNING, 0,
                                GD_MSG_VOLINFO_SET_FAIL,
                                "failed to store volinfo");
                        goto out;
                }

                ret = 0;
                goto out;
        }

        case GF_OP_CMD_DETACH_START:
        case GF_OP_CMD_START:
                /* Reset defrag status to 'NOT STARTED' whenever a
                 * remove-brick/rebalance command is issued to remove
                 * stale information from previous run.
                 */
                volinfo->rebal.defrag_status = GF_DEFRAG_STATUS_NOT_STARTED;
                ret = dict_get_str (dict, GF_REMOVE_BRICK_TID_KEY, &task_id_str);
                if (ret) {
                        gf_msg_debug (this->name, errno,
                                "Missing remove-brick-id");
                        ret = 0;
                } else {
                        gf_uuid_parse (task_id_str, volinfo->rebal.rebalance_id) ;
                        volinfo->rebal.op = GD_OP_REMOVE_BRICK;
                }
                force = 0;
                break;

        case GF_OP_CMD_COMMIT:
                force = 1;
                break;

        case GF_OP_CMD_DETACH_COMMIT:
        case GF_OP_CMD_DETACH_COMMIT_FORCE:
                glusterd_op_perform_detach_tier (volinfo);
                detach_commit = 1;

                /* Disabling ctr when detaching a tier, since
                 * currently tier is the only consumer of ctr.
                 * Revisit this code when this constraint no
                 * longer exist.
                 */
                dict_del (volinfo->dict, "features.ctr-enabled");
                dict_del (volinfo->dict, "cluster.tier-mode");

                hot_shd_key = gd_get_shd_key (volinfo->tier_info.hot_type);
                cold_shd_key = gd_get_shd_key (volinfo->tier_info.cold_type);
                if (hot_shd_key) {
                        /*
                         * Since post detach, shd graph will not contain hot
                         * tier. So we need to clear option set for hot tier.
                         * For a tiered volume there can be different key
                         * for both hot and cold. If hot tier is shd compatible
                         * then we need to remove the configured value when
                         * detaching a tier, only if the key's are different or
                         * cold key is NULL. So we will set delete_key first,
                         * and if cold key is not null and they are equal then
                         * we will clear the flag. Otherwise we will delete the
                         * key.
                         */
                        if (cold_shd_key)
                                delete_key = strcmp (hot_shd_key, cold_shd_key);
                        if (delete_key)
                                dict_del (volinfo->dict, hot_shd_key);
               }
                /* fall through */

        case GF_OP_CMD_COMMIT_FORCE:

                if (volinfo->decommission_in_progress) {
                        if (volinfo->rebal.defrag) {
                                LOCK (&volinfo->rebal.defrag->lock);
                                /* Fake 'rebalance-complete' so the graph change
                                   happens right away */
                                volinfo->rebal.defrag_status =
                                                GF_DEFRAG_STATUS_COMPLETE;

                                UNLOCK (&volinfo->rebal.defrag->lock);
                        }
                        /* Graph change happens in rebalance _cbk function,
                           no need to do anything here */
                        /* TODO: '_cbk' function is not doing anything for now */
                }

                ret = 0;
                force = 1;
                break;
        }

        ret = dict_get_int32 (dict, "count", &count);
        if (ret) {
                gf_msg ("glusterd", GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "Unable to get count");
                goto out;
        }

        if (volinfo->type == GF_CLUSTER_TYPE_TIER)
                count = glusterd_set_detach_bricks(dict, volinfo);

        /* Save the list of bricks for later usage only on starting a
         * remove-brick. Right now this is required for displaying the task
         * parameters with task status in volume status.
         */

        if (start_remove) {
                bricks_dict = dict_new ();
                if (!bricks_dict) {
                        ret = -1;
                        goto out;
                }
                ret = dict_set_int32 (bricks_dict, "count", count);
                if (ret) {
                        gf_msg (this->name, GF_LOG_ERROR, errno,
                                GD_MSG_DICT_SET_FAILED,
                                "Failed to save remove-brick count");
                        goto out;
                }
        }

        while ( i <= count) {
                snprintf (key, 256, "brick%d", i);
                ret = dict_get_str (dict, key, &brick);
                if (ret) {
                        gf_msg (this->name, GF_LOG_ERROR, errno,
                                GD_MSG_DICT_GET_FAILED, "Unable to get %s",
                                key);
                        goto out;
                }

                if (start_remove) {
                        brick_tmpstr = gf_strdup (brick);
                        if (!brick_tmpstr) {
                                ret = -1;
                                gf_msg (this->name, GF_LOG_ERROR, ENOMEM,
                                        GD_MSG_NO_MEMORY,
                                        "Failed to duplicate brick name");
                                goto out;
                        }
                        ret = dict_set_dynstr (bricks_dict, key, brick_tmpstr);
                        if (ret) {
                                gf_msg (this->name, GF_LOG_ERROR, errno,
                                        GD_MSG_DICT_SET_FAILED,
                                        "Failed to add brick to dict");
                                goto out;
                        }
                        brick_tmpstr = NULL;
                }

                ret = glusterd_op_perform_remove_brick (volinfo, brick, force,
                                                        &need_rebalance);
                if (ret)
                        goto out;
                i++;
        }

        if (detach_commit) {
                 /* Clear related information from volinfo */
                tier_info = ((void *)(&volinfo->tier_info));
                memset (tier_info, 0, sizeof (volinfo->tier_info));
        }

        if (start_remove)
                volinfo->rebal.dict = dict_ref (bricks_dict);

        ret = dict_get_int32 (dict, "replica-count", &replica_count);
        if (!ret) {
                gf_msg (this->name, GF_LOG_INFO, errno,
                        GD_MSG_DICT_GET_FAILED,
                        "changing replica count %d to %d on volume %s",
                        volinfo->replica_count, replica_count,
                        volinfo->volname);
                volinfo->replica_count = replica_count;
                /* A reduction in replica count implies an arbiter volume
                 * earlier is now no longer one. */
                if (volinfo->arbiter_count)
                        volinfo->arbiter_count = 0;
                volinfo->sub_count = replica_count;
                volinfo->dist_leaf_count = glusterd_get_dist_leaf_count (volinfo);

                /*
                 * volinfo->type and sub_count have already been set for
                 * volumes undergoing a detach operation, they should not
                 * be modified here.
                 */
                if ((replica_count == 1) && (cmd != GF_OP_CMD_DETACH_COMMIT) &&
                    (cmd != GF_OP_CMD_DETACH_COMMIT_FORCE)) {
                        if (volinfo->type == GF_CLUSTER_TYPE_REPLICATE) {
                                volinfo->type = GF_CLUSTER_TYPE_NONE;
                                /* backward compatibility */
                                volinfo->sub_count = 0;
                        } else {
                                volinfo->type = GF_CLUSTER_TYPE_STRIPE;
                                /* backward compatibility */
                                volinfo->sub_count = volinfo->dist_leaf_count;
                        }
                }
        }
        volinfo->subvol_count = (volinfo->brick_count /
                                 volinfo->dist_leaf_count);

        ret = glusterd_create_volfiles_and_notify_services (volinfo);
        if (ret) {
                gf_msg (this->name, GF_LOG_WARNING, 0,
                        GD_MSG_VOLFILE_CREATE_FAIL, "failed to create volfiles");
                goto out;
        }

        ret = glusterd_store_volinfo (volinfo, GLUSTERD_VOLINFO_VER_AC_INCREMENT);
        if (ret) {
                gf_msg (this->name, GF_LOG_WARNING, 0,
                        GD_MSG_VOLINFO_STORE_FAIL, "failed to store volinfo");
                goto out;
        }

        if (start_remove &&
            volinfo->status == GLUSTERD_STATUS_STARTED) {
                ret = glusterd_svcs_reconfigure ();
                if (ret) {
                        gf_msg (this->name, GF_LOG_WARNING, 0,
                                GD_MSG_NFS_RECONF_FAIL,
                               "Unable to reconfigure NFS-Server");
                goto out;
                }
        }

        /* Need to reset the defrag/rebalance status accordingly */
        switch (volinfo->rebal.defrag_status) {
        case GF_DEFRAG_STATUS_FAILED:
        case GF_DEFRAG_STATUS_COMPLETE:
                volinfo->rebal.defrag_status = 0;
        default:
                break;
        }
        if (!force && need_rebalance) {
                if (dict_get_uint32(dict, "commit-hash", &commit_hash) == 0) {
                        volinfo->rebal.commit_hash = commit_hash;
                }
                /* perform the rebalance operations */
                defrag_cmd = GF_DEFRAG_CMD_START_FORCE;
                if (cmd == GF_OP_CMD_DETACH_START)
                        defrag_cmd = GF_DEFRAG_CMD_START_DETACH_TIER;
                ret = glusterd_handle_defrag_start
                        (volinfo, err_str, sizeof (err_str),
                         defrag_cmd,
                         glusterd_remove_brick_migrate_cbk, GD_OP_REMOVE_BRICK);

                if (!ret)
                        volinfo->decommission_in_progress = 1;

                if (ret) {
                        gf_msg (this->name, GF_LOG_ERROR, 0,
                                GD_MSG_REBALANCE_START_FAIL,
                                "failed to start the rebalance");
                }
        } else {
                if (GLUSTERD_STATUS_STARTED == volinfo->status)
                        ret = glusterd_svcs_manager (volinfo);
        }
out:
       if (ret && err_str[0] && op_errstr)
                *op_errstr = gf_strdup (err_str);

        GF_FREE (brick_tmpstr);
        if (bricks_dict)
                dict_unref (bricks_dict);

        return ret;
}

int
glusterd_op_stage_barrier (dict_t *dict, char **op_errstr)
{
        int                  ret         = -1;
        xlator_t             *this       = NULL;
        char                 *volname    = NULL;
        glusterd_volinfo_t   *vol        = NULL;

        GF_ASSERT (dict);
        this = THIS;
        GF_ASSERT (this);

        ret = dict_get_str (dict, "volname", &volname);
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "Volname not present in "
                        "dict");
                goto out;
        }

        ret = glusterd_volinfo_find (volname, &vol);
        if (ret) {
                gf_asprintf (op_errstr, "Volume %s does not exist", volname);
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_VOL_NOT_FOUND, "%s", *op_errstr);
                goto out;
        }

        if (!glusterd_is_volume_started (vol)) {
                gf_asprintf (op_errstr, "Volume %s is not started", volname);
                ret = -1;
                goto out;
        }

        ret = dict_get_str_boolean (dict, "barrier", -1);
        if (ret == -1) {
                gf_asprintf (op_errstr, "Barrier op for volume %s not present "
                             "in dict", volname);
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "%s", *op_errstr);
                goto out;
        }
        ret = 0;
out:
        gf_msg_debug (this->name, 0, "Returning %d", ret);
        return ret;
}

int
glusterd_op_barrier (dict_t *dict, char **op_errstr)
{
        int                  ret         = -1;
        xlator_t             *this       = NULL;
        char                 *volname    = NULL;
        glusterd_volinfo_t   *vol        = NULL;
        char                 *barrier_op = NULL;

        GF_ASSERT (dict);
        this = THIS;
        GF_ASSERT (this);

        ret = dict_get_str (dict, "volname", &volname);
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "Volname not present in "
                        "dict");
                goto out;
        }

        ret = glusterd_volinfo_find (volname, &vol);
        if (ret) {
                gf_asprintf (op_errstr, "Volume %s does not exist", volname);
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_VOL_NOT_FOUND, "%s", *op_errstr);
                goto out;
        }

        ret = dict_get_str (dict, "barrier", &barrier_op);
        if (ret) {
                gf_asprintf (op_errstr, "Barrier op for volume %s not present "
                             "in dict", volname);
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_GET_FAILED, "%s", *op_errstr);
                goto out;
        }

        ret = dict_set_dynstr_with_alloc (vol->dict, "features.barrier",
                                          barrier_op);
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, errno,
                        GD_MSG_DICT_SET_FAILED, "Failed to set barrier op in"
                        " volume option dict");
                goto out;
        }

        gd_update_volume_op_versions (vol);
        ret = glusterd_create_volfiles (vol);
        if (ret) {
                gf_msg (this->name, GF_LOG_ERROR, 0,
                        GD_MSG_VOLFILE_CREATE_FAIL, "Failed to create volfiles");
                goto out;
        }
        ret = glusterd_store_volinfo (vol, GLUSTERD_VOLINFO_VER_AC_INCREMENT);

out:
        gf_msg_debug (this->name, 0, "Returning %d", ret);
        return ret;
}

int
glusterd_handle_attach_tier (rpcsvc_request_t *req)
{
        return glusterd_big_locked_handler (req, __glusterd_handle_add_brick);
}

int
glusterd_handle_detach_tier (rpcsvc_request_t *req)
{
        return glusterd_big_locked_handler (req,
                                            __glusterd_handle_remove_brick);
}