IV36074: JFS2 FILESYSTEM HANG IN BMASSIGN

APAR status

• Closed as program error.

Error description

• On AIX 6.1 TL5 and later TLs a JFS2 filesystem may hang,
  causing all commands accessing that filesystem to hang
  and to be unkillable.  The only way to recover is to
  reboot the system.
  
  The dump will show a thread waiting in bmAssign:
  [000D50D0]e_block_thread+000290 ()
  [000D5D28]e_sleep_thread+0000E8 (??, ??, ??)
  [00260B38]bmAssign+000778 (??, ??, ??, ??, ??, ??)
  [0025FDFC]bmRead+00009C (??, ??, ??, ??, ??, ??)
  [002B363C]xtTruncate+0009DC (??, ??, ??, ??, ??)
  [00299FDC]txLog+00029C (??, ??, ??)
  [0029C3D4]txCommit+000654 (??, ??, ??, ??)
  [002D3C58]j2_remove+000438 (??, ??, ??, ??)
  [005838A4]vnop_remove+0003E4 (??, ??, ??, ??)
  [0067B9C0]kunlink+000300 (??, ??)
  and other threads waiting for locks held by that thread
  or waiting in bmFlushRead for the same pager buffer.
  

Local fix

• Reboot the system.
  

Problem summary

• hand on bmAssign with following stack:
  
  pvthread+0B6E00 STACK:
   000D3870 e_block_thread+000290 ()
   000D44C8 e_sleep_thread+0000E8 (??, ??, ??)
   0023D258 bmAssign+000778 (??, ??, ??, ??, ??, ??)
   0023C51C bmRead+00009C (??, ??, ??, ??, ??, ??)
   00289F8C xtFreeMap+00018C (??, ??, ??)
   00275094 txUpdateMap+000254 (??, ??)
   00278258 txCommit+0006D8 (??, ??, ??, ??)
   002AF178 j2_remove+000438 (??, ??, ??, ??)
   0057C0A4 vnop_remove+0003E4 (??, ??, ??, ??)
   00672580 kunlink+000300 (??, ??)
   00003850 ovlya_addr_sc_flih_main+000130 ()
   kdb_get_virtual_memory  no real storage @ 700000000596440
   1000624B4 00000001000624B4 ()
  

Problem conclusion

• Track buffer RDONLY users via separate data field and manage
  accordingly
  

Temporary fix

Comments

APAR Information

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

  IV16158

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

Fix information

• Fixed component name

  AIX V7.1

• Fixed component ID

  5765H4000

Applicable component levels

• R710 PSY

     UP