mongodb - Mongod 在启动时死于致命断言 16252

Question

MongoDB 在我构建的 VM 上继续崩溃。用谷歌搜索致命断言编号无济于事。任何指针将不胜感激...

详细信息:

在运行 Ubuntu 14.04 的虚拟机系统上。来自 10gen 的最新稳定 mongo 包:

# dpkg -l | grep mongo
ii  libmongo-client0:amd64                     0.1.7-1                             amd64        Alternate C driver for the MongoDB document-oriented datastore
ii  mongodb-org                                2.6.3                               amd64        MongoDB open source document-oriented database system (metapackage)
ii  mongodb-org-mongos                         2.6.3                               amd64        MongoDB sharded cluster query router
ii  mongodb-org-server                         2.6.3                               amd64        MongoDB database server
ii  mongodb-org-shell                          2.6.3                               amd64        MongoDB shell client
ii  mongodb-org-tools                          2.6.3                               amd64        MongoDB tools

大量可用磁盘。

来自/var/log/mongodb/mongod.log 的日志:

2014-06-30T16:11:14.474-0600 ***** SERVER RESTARTED *****
2014-06-30T16:11:14.479-0600 [initandlisten] MongoDB starting : pid=27406 port=27017 dbpath=/var/lib/mongodb 64-bit host=foobar
2014-06-30T16:11:14.479-0600 [initandlisten] db version v2.6.3
2014-06-30T16:11:14.479-0600 [initandlisten] git version: 255f67a66f9603c59380b2a389e386910bbb52cb
2014-06-30T16:11:14.479-0600 [initandlisten] build info: Linux build12.nj1.10gen.cc 2.6.32-431.3.1.el6.x86_64 #1 SMP Fri Jan 3 21:39:27 UTC 2014 x86_64 BOOST_LIB_VERSION=1_49
2014-06-30T16:11:14.479-0600 [initandlisten] allocator: tcmalloc
2014-06-30T16:11:14.479-0600 [initandlisten] options: { config: "/etc/mongod.conf", net: { bindIp: "127.0.0.1" }, storage: { dbPath: "/var/lib/mongodb" }, systemLog: { destination: "file", logAppend: true, path: "/var/log/mongodb/mongod.log" } }
2014-06-30T16:11:14.481-0600 [initandlisten] journal dir=/var/lib/mongodb/journal
2014-06-30T16:11:14.482-0600 [initandlisten] recover begin
2014-06-30T16:11:14.482-0600 [initandlisten] info no lsn file in journal/ directory
2014-06-30T16:11:14.482-0600 [initandlisten] recover lsn: 0
2014-06-30T16:11:14.482-0600 [initandlisten] recover /var/lib/mongodb/journal/j._0
2014-06-30T16:11:14.498-0600 [initandlisten] recover cleaning up
2014-06-30T16:11:14.498-0600 [initandlisten] removeJournalFiles
2014-06-30T16:11:14.500-0600 [initandlisten] recover done
2014-06-30T16:11:14.720-0600 [initandlisten] waiting for connections on port 27017
2014-06-30T16:11:14.751-0600 [IndexRebuilder] Fatal Assertion 16252
2014-06-30T16:11:14.758-0600 [IndexRebuilder] 0x11da801 0x117c9f9 0x115f43d 0xa1a7cf 0xa1ccab 0xa1ced3 0x8f4ad0 0xb89da3 0xb8ac78 0x11625f2 0x121f219 0x7fad84a56182 0x7fad83d5b30d
 /usr/bin/mongod(_ZN5mongo15printStackTraceERSo+0x21) [0x11da801]
 /usr/bin/mongod(_ZN5mongo10logContextEPKc+0x159) [0x117c9f9]
 /usr/bin/mongod(_ZN5mongo13fassertFailedEi+0xcd) [0x115f43d]
 /usr/bin/mongod(_ZN5mongo4Lock7DBWrite9lockOtherERKNS_10StringDataE+0xa6f) [0xa1a7cf]
 /usr/bin/mongod(_ZN5mongo4Lock7DBWrite6lockDBERKSs+0x17b) [0xa1ccab]
 /usr/bin/mongod(_ZN5mongo4Lock7DBWriteC1ERKNS_10StringDataE+0x63) [0xa1ced3]
 /usr/bin/mongod(_ZN5mongo6Client12WriteContextC1ERKSsS3_+0x40) [0x8f4ad0]
 /usr/bin/mongod(_ZN5mongo14IndexRebuilder7checkNSERKSt4listISsSaISsEE+0xd3) [0xb89da3]
 /usr/bin/mongod(_ZN5mongo14IndexRebuilder3runEv+0x188) [0xb8ac78]
 /usr/bin/mongod(_ZN5mongo13BackgroundJob7jobBodyEv+0xd2) [0x11625f2]
 /usr/bin/mongod() [0x121f219]
 /lib/x86_64-linux-gnu/libpthread.so.0(+0x8182) [0x7fad84a56182]
 /lib/x86_64-linux-gnu/libc.so.6(clone+0x6d) [0x7fad83d5b30d]
2014-06-30T16:11:14.758-0600 [IndexRebuilder]

***aborting after fassert() failure

Answer 1

看起来像是损坏的日志文件或整个数据库(不确定这是怎么发生的？)。移动数据库文件，并从备份转储中恢复，一切看起来都很好。

这在 apt-get 升级后再次发生。想知道安装脚本是否在升级前没有正确关闭数据库，从而破坏了日志？