Kbase P120002: Batch client disconnected in promon don't disconnect properly.
Autor |
  Progress Software Corporation - Progress |
Acesso |
  Público |
Publicação |
  08/08/2008 |
|
Status: Unverified
SYMPTOM(s):
Batch client disconnected in promon don't disconnect properly.
Batch processes continue to gain time at OS level but nothing changes in promon for the records locked by the process.
Batch processes holding locks in database as seen from promon:
Usr Name Chain # Rec-id Table Lock Flags Tran State Tran ID
215 batch REC 4802 12241531 142 SHR Dead 203522 5023
215 batch REC 6490 3337454 18 SHR L Active 203522 5377
215 batch REC 6989 3467602 18 EXCL Active 203522 5377
215 batch REC 9030 12305563 176 SHR Dead 0
215 batch TAB 9991 0 18 IX L Active 203522 5377
Stack trace generated by killing the process contains the following:
9000000000b0d28 select (?, ?, ?, ?, ?) + ??
1003bb6b4 rnSocketSelect (?, ?) + 1e4
1003b6d24 wvUnixGetKey (?) + 130
1003b8804 wvGetEvent (?, ?) + d4
100569cec wv_get_alert_event (?) + 34
10056a2d0 wvAlertBox (?, ?) + 64c
FACT(s) (Environment):
UNIX
Progress/OpenEdge Versions
CAUSE:
Network error caused the telnet session to be severed from the machine leaving the client process started from that telnet session unaware that the session was stranded.
FIX:
Issuing kill -s HUP to the stranded client session caused the session to back out of incomplete work and terminate normally.