IJ48629: LOGASSERTFAILED: INODE.FILESIZE.GETVAL() == TINODEP->FILESIZE

APAR status

• Closed as program error.

Error description

• GPFS crash due to the following Assertion failure.
  
  2023-07-31_15:59:22.722-0500: [X] logAssertFailed:
  inode.fileSize.getval() == tinodeP->fileSize
  2023-07-31_15:59:22.722-0500: [X] return code 3968,
  reason code 3744, log record tag 0
  2023-07-31_15:59:23.900-0500: [X] *** Assert
  exp(inode.fileSize.getval() == tinodeP->fileSize) in line
  5722 of file
  /project/sprelgpfs516/build/rgpfs516s001d/src/avs/fs/mmf
  s/ts/fs/metadata.C
  2023-07-31_15:59:23.900-0500: [E] *** Traceback:
  2023-07-31_15:59:23.900-0500: [E] 2:0x5557D3F5136A
  logAssertFailed + 0x3AA at ??:0
  2023-07-31_15:59:23.900-0500: [E] 3:0x5557D3C883E4
  FileMetadata::mergeInode(int) + 0xF74 at ??:0
  2023-07-31_15:59:23.900-0500: [E] 4:0x5557D3D23DFA
  FileMetadata::mnbMergeInode(int, Inode*) + 0x12A at ??:0
  2023-07-31_15:59:23.900-0500: [E] 5:0x5557D3D2C791
  FileMetadata::MetaNodeBegin(int, Inode*) + 0x291 at ??:0
  2023-07-31_15:59:23.900-0500: [E] 6:0x5557D3D1C828
  MNodeToken::tryToBecomeMnode(LkObj::LockModeEnum, int,
  Inode*) + 0x588 at ??:0
  2023-07-31_15:59:23.900-0500: [E] 7:0x5557D3D24DA6
  InodeLkObj::change_token(CacheObj*, LkObj*,
  LkObj::LockModeEnum, int, LkObj::LockModeEnum*) + 0x156
  at ??:0
  2023-07-31_15:59:23.900-0500: [E] 8:0x5557D3A4354F
  LkObj::change_lock_shark_m(CacheObj*,
  LkObj::LockModeEnum, LkObj::LockModeEnum,
  LkObj::LockModeEnum*, int, int) +
  0x4EF at ??:0
  2023-07-31_15:59:23.900-0500: [E] 9:0x5557D3A440D5
  LkObj::lock_shark_m(CacheObj*, LkObj::LockModeEnum,
  LkObj::LockModeEnum*, int, int) + 0x15 at ??:0
  2023-07-31_15:59:23.900-0500: [E] 10:0x5557D3CAE158
  LockFile(OpenFile**, StripeGroup*, FileUID,
  OperationLockMode, LkObj::LockModeEnum,
  LkObj::LockModeEnum*, LkObj:
  :LockModeEnum*, int, int) + 0x2B8 at ??:0
  2023-07-31_15:59:23.900-0500: [E] 11:0x5557D3BFF47B
  FSOperation::createLockedFile(StripeGroup*, FileUID,
  OperationLockMode, LkObj::LockModeEnum, OpenFile**,
  unsigne
  d int*, int, int) + 0x9B at ??:0
  2023-07-31_15:59:23.900-0500: [E] 12:0x5557D3B247A9
  InodePrefetchInstance::doWork() + 0x769 at ??:0
  2023-07-31_15:59:23.900-0500: [E] 13:0x5557D3B24E48
  InodePrefetchInstance::WorkerThreadBody(void*) + 0x108 at
  ??:0
  2023-07-31_15:59:23.900-0500: [E] 14:0x5557D3A1F662
  Thread::callBody(Thread*) + 0x42 at ??:0
  2023-07-31_15:59:23.900-0500: [E] 15:0x5557D3A0C680
  Thread::callBodyWrapper(Thread*) + 0xA0 at ??:0
  2023-07-31_15:59:23.900-0500: [E] 16:0x7EFCD480A1CA
  start_thread + 0xEA at ??:0
  2023-07-31_15:59:23.900-0500: [E] 17:0x7EFCD3529E73
  __GI___clone + 0x43 at ??:0
  

Local fix

• GPFS daemon will restart itself. It's a code defect. One
  needs to get the fix code.
  

Problem summary

• Race between stat/gpfs_stalite() and inode token revoke causes
  log assert.
  

Problem conclusion

• This problem is fixed in 5.1.2.14
  To see all Spectrum Scale APARs and their respective
  Fix solutions refer to page:
  https://public.dhe.ibm.com/storage/spectrumscale/spectrum_scale_
  apars.html
  
  Benefits of the solution:
  Closed race window and fixed this problem
  
  Work Around:
  Set config parameters statliteMaxAttrAge and statMaxAttrAge to 0
  to disable stat lite.
  
  Problem trigger:
  A file is actively written on one node and stat() or
  gpfs_statlite() is called repeatedly on another node
  
  Symptom:
  Abend/Crash
  
  Platforms affected:
  ALL Operating System environments
  
  Functional Area affected:
  All Scale Users
  
  Customer Impact:
  High Importance
  

Temporary fix

Comments

APAR Information

• APAR is sysrouted FROM one or more of the following:

• APAR is sysrouted TO one or more of the following:

Fix information

• Fixed component name

  SPEC SCALE STD

• Fixed component ID

  5737F33AP

Applicable component levels