Percona MongoDB does not start after Azure VM restart...

[Andre Esteves]

Hi everyone,

Last week after a restart of a Azure VM my Percona MongoDB working with RocksDB engine stopped working properly...

I tried with --repair but the result is the same. Can anyone help?

2017-08-20T22:03:11.997+0000 I CONTROL  [main] ***** SERVER RESTARTED *****

2017-08-20T22:03:12.005+0000 I STORAGE  [main] [RocksDB] Block Cache Size GB: 0

2017-08-20T22:03:12.005+0000 I STORAGE  [main] [RocksDB] Compression: snappy

2017-08-20T22:03:12.005+0000 I STORAGE  [main] [RocksDB] MaxWriteMBPerSec: 1024

2017-08-20T22:03:12.005+0000 I STORAGE  [main] [RocksDB] Engine custom option: 

2017-08-20T22:03:12.005+0000 I STORAGE  [main] [RocksDB] Crash safe counters: 0

2017-08-20T22:03:12.005+0000 I STORAGE  [main] [RocksDB] Counters: 0

2017-08-20T22:03:12.005+0000 I STORAGE  [main] [RocksDB] Use SingleDelete in index: 0

2017-08-20T22:03:12.008+0000 I ACCESS   [main] Initialized External Auth Session

2017-08-20T22:03:12.015+0000 I CONTROL  [initandlisten] MongoDB starting : pid=20429 port=27017 dbpath=/datadrive/mongodb 64-bit host=hiddenOnPurpose

2017-08-20T22:03:12.015+0000 I CONTROL  [initandlisten] db version v3.2.15-3.5

2017-08-20T22:03:12.015+0000 I CONTROL  [initandlisten] git version: 3615e6f7425b87278261a9f2d3e52a14ff3c2a46

2017-08-20T22:03:12.015+0000 I CONTROL  [initandlisten] OpenSSL version: OpenSSL 1.0.2g  1 Mar 2016

2017-08-20T22:03:12.015+0000 I CONTROL  [initandlisten] allocator: tcmalloc

2017-08-20T22:03:12.015+0000 I CONTROL  [initandlisten] modules: none

2017-08-20T22:03:12.015+0000 I CONTROL  [initandlisten] build environment:

2017-08-20T22:03:12.015+0000 I CONTROL  [initandlisten]     distarch: x86_64

2017-08-20T22:03:12.015+0000 I CONTROL  [initandlisten]     target_arch: x86_64

2017-08-20T22:03:12.015+0000 I CONTROL  [initandlisten] options: { config: "/etc/mongod.conf", net: { bindIp: true, port: 27017 }, processManagement: { fork: true, pidFilePath: "/

var/run/mongod.pid" }, security: { authorization: "enabled" }, setParameter: { failIndexKeyTooLong: "false", internalQueryExecYieldIterations: "100000", internalQueryExecYieldPeri

odMS: "1000" }, storage: { dbPath: "/datadrive/mongodb", engine: "rocksdb", journal: { enabled: true } }, systemLog: { destination: "file", logAppend: true, path: "/datadrive/log/

mongod.log" } }

2017-08-20T22:03:12.067+0000 W -        [initandlisten] Detected unclean shutdown - /datadrive/mongodb/mongod.lock is not empty.

2017-08-20T22:03:12.293+0000 I STORAGE  [initandlisten] 126 dropped prefixes need compaction

2017-08-20T22:03:12.332+0000 I STORAGE  [initandlisten] dropping unused ident: index-0--421938426481405637

2017-08-20T22:03:12.381+0000 I STORAGE  [initandlisten] dropping unused ident: index-1--421938426481405637

2017-08-20T22:03:12.496+0000 I -        [initandlisten] Invariant failure prefixIter != _identPrefixMap.end() src/mongo/db/storage/rocks/src/rocks_engine.cpp 501

2017-08-20T22:03:12.496+0000 I -        [initandlisten] 

***aborting after invariant() failure

2017-08-20T22:03:12.535+0000 F -        [initandlisten] Got signal: 6 (Aborted).

Thank you very much in advance,

Andre

[Denis Protivenskii]

Hi,

Please redirect your question to more appropriate group: https://groups.google.com/forum/?utm_medium=email&utm_source=footer#!forum/mongo-rocks
Also, provide info on your OS and filesystem.

- Denis.