MongoDB不可恢复故障

我的MacOSX9有一个电源故障，当重启时，我在Mongo中有一个完全无法恢复的db。我使用dbpath和logpath指定选项的db per目录，并且能够恢复除一个dbs之外的所有dbs。这是不可恢复日志中的日志。任何想法都会有帮助-我已经运行了-修复并删除了锁定文件，结果为零

日志文件：

2014-12-01T09:40:40.607-0700 [initandlisten] MongoDB starting : pid=1273 port=27017 dbpath=/mongo/dbs/ 64-bit host=bbcoms-imac-2.body.local
2014-12-01T09:40:40.607-0700 [initandlisten]
2014-12-01T09:40:40.607-0700 [initandlisten] ** WARNING: soft rlimits too low. Number of files is 256, should be at least 1000
2014-12-01T09:40:40.608-0700 [initandlisten] db version v2.6.5
2014-12-01T09:40:40.608-0700 [initandlisten] git version: nogitversion
2014-12-01T09:40:40.608-0700 [initandlisten] build info: Darwin minimavericks.local 13.4.0 Darwin Kernel Version 13.4.0: Sun Aug 17 19:50:11 PDT 2014; root:xnu-2422.115.4~1/RELEASE_X86_64 x86_64 BOOST_LIB_VERSION=1_49
2014-12-01T09:40:40.608-0700 [initandlisten] allocator: tcmalloc
2014-12-01T09:40:40.608-0700 [initandlisten] options: { repair: true, storage: { dbPath: "/mongo/dbs/", repairPath: "/mongo/db" } }
2014-12-01T09:40:40.608-0700 [initandlisten] exception in initAndListen: 12590 repairpath (/mongo/db) does not exist, terminating
2014-12-01T09:40:40.608-0700 [initandlisten] dbexit:
2014-12-01T09:40:40.608-0700 [initandlisten] shutdown: going to close listening sockets...
2014-12-01T09:40:40.608-0700 [initandlisten] shutdown: going to flush diaglog...
2014-12-01T09:40:40.609-0700 [initandlisten] shutdown: going to close sockets...
2014-12-01T09:20:58.339-0700 [clientcursormon]  mapped (incl journal view):2176
2014-12-01T09:20:58.339-0700 [clientcursormon]  connections:1
2014-12-01T09:25:58.390-0700 [clientcursormon] mem (MB) res:41 virt:4655
2014-12-01T09:25:58.390-0700 [clientcursormon]  mapped (incl journal view):2176
2014-12-01T09:25:58.390-0700 [clientcursormon]  connections:1
2014-12-01T09:28:53.518-0700 [conn1] end connection 127.0.0.1:50225 (0 connections now open)
2014-12-01T09:30:55.153-0700 [initandlisten] connection accepted from 127.0.0.1:50411 #3 (1 connection now open)
2014-12-01T09:30:57.586-0700 [conn3] command admin.$cmd command: listDatabases { listDatabases: 1.0 } keyUpdates:0 numYields:0 locks(micros) R:3 W:105826 r:14 reslen:337 106ms
2014-12-01T09:30:58.313-0700 [TTLMonitor] articles.system.indexes Assertion failure isOk() src/mongo/db/storage/extent.h 80
2014-12-01T09:30:58.316-0700 [TTLMonitor] articles.system.indexes 0x1010665aa 0x101011135 0x100ffdbf2 0x100e1ca0c 0x100e4d4bc 0x100977b13 0x100ad7187 0x100cfdbcb 0x100ce5e6d 0x100b7d0b7 0x100b85055 0x100b852bb 0x1008dae1b 0x1008bee8a 0x100b854f1 0x100e50be6 0x100e4fe07 0x101000606 0x10109f2a1 0x7fff84fd0899
 0   mongod                              0x00000001010665aa _ZN5mongo15printStackTraceERNSt3__113basic_ostreamIcNS0_11char_traitsIcEEEE + 58
 1   mongod                              0x0000000101011135 _ZN5mongo10logContextEPKc + 453
 2   mongod                              0x0000000100ffdbf2 _ZN5mongo12verifyFailedEPKcS1_j + 626
 3   mongod                              0x0000000100e1ca0c _ZNK5mongo13ExtentManager9getExtentERKNS_7DiskLocEb + 108
 4   mongod                              0x0000000100e4d4bc _ZN5mongo12FlatIteratorC2EPKNS_10CollectionERKNS_7DiskLocERKNS_20CollectionScanParams9DirectionE + 92
 5   mongod                              0x0000000100977b13 _ZNK5mongo10Collection11getIteratorERKNS_7DiskLocEbRKNS_20CollectionScanParams9DirectionE + 115
 6   mongod                              0x0000000100ad7187 _ZN5mongo14CollectionScan4workEPm + 519
 7   mongod                              0x0000000100cfdbcb _ZN5mongo12PlanExecutor7getNextEPNS_7BSONObjEPNS_7DiskLocE + 283
 8   mongod                              0x0000000100ce5e6d _ZN5mongo11newRunQueryERNS_7MessageERNS_12QueryMessageERNS_5CurOpES1_ + 3101
 9   mongod                              0x0000000100b7d0b7 _ZN5mongo16assembleResponseERNS_7MessageERNS_10DbResponseERKNS_11HostAndPortE + 2663
 10  mongod                              0x0000000100b85055 _ZN5mongo14DBDirectClient4callERNS_7MessageES2_bPNSt3__112basic_stringIcNS3_11char_traitsIcEENS3_9allocatorIcEEEE + 165
2014-12-01T09:40:40.609-0700 [initandlisten] shutdown: waiting for fs preallocator...
2014-12-01T09:40:40.609-0700 [initandlisten] shutdown: closing all files...
2014-12-01T09:40:40.610-0700 [initandlisten] closeAllFiles() finished
2014-12-01T09:40:40.610-0700 [initandlisten] dbexit: really exiting now

---

我看到了一个带有区段的断言失败的错误，但是上面有行

2014-12-01T09:40:40.608-0700 [initandlisten] options: { repair: true, storage: { dbPath: "/mongo/dbs/", repairPath: "/mongo/db" } }
2014-12-01T09:40:40.608-0700 [initandlisten] exception in initAndListen: 12590 repairpath (/mongo/db) does not exist, terminating

---

因此，在运行修复时，似乎没有提供有效的修复路径。你能启动服务器吗？这个mongod是副本集的一部分吗？如果服务器可以启动，您可以在坏数据库中的集合上运行吗？发生了什么事？

谢谢，但我无法启动服务器。我最终不得不把那个特定的数据库吹走才能开始。这不是最好的解决方案，但我找不到任何允许使用该数据进行操作的方法。有点可怕。两年后这里又有一个评论，但我最近也遇到了类似的问题。您可以通过将数据文件直接移动到新实例来恢复它们。将*.0、*.1等和*.ns文件移动到新的db目录路径中，然后启动mongo，它将拾取它们。如果使用配置设置：directoryPerDB:true，它将提供选择要恢复的数据库的功能。请记住，数据库路径位于存储部分的dbPath下的配置文件中。