I have recently seen in my system that all update work process hanged and no was able able to work and the wp process started timing out.
here is the mail part of wp trace
M Mon Jan 25 15:44:15 2010
M ThRollOut: act roll state = DP_ROLLED_OUT
M ThRollOut: roll level <> 0, don't call ab_rollout
M ThRollOut: full roll out of U22 M0 I0 (Level 7) ok
M ThIErrHandle: prv_action of W21: 0x8
M ThCallDbBreak: use db_sqlbreak
C SQLBREAK: DBSL_CMD_SQLBREAK: CbOnCancel was not set. rc: 15
M *** WARNING => sql break failed (16388)
M ThIErrHandle: don't try rollback again
M ThIErrHandle: do not call ThrCoreInfo (no_core_info=0, in_dynp_env=1)
M ThIErrHandle: Entering ThReplyToMsg
M ThIErrHandle: Entering ThErrHdlUser
M ThErrHdlUser: set th_errno (11)
M ThErrHdlUser: save context
M ThEmContextDetach2: detach T21/M0 from em memory (em_hdl=60, force=0)
M ThEmContextDetach2: reset local em info
M call ThrShutDown ...
B Disconnecting from ALL connections:
B Wp Hdl ConName ConId ConState TX PRM RCT TIM MAX OPT Date Time DBHost
B 021 000 R/3 000000000 ACTIVE YES YES NO 000 255 255 20091218 191945 ERPSQL
C DbSlDisconnect NOT DONE because thread IDs are different.
B Disconnected from connection 0
B statistics db_con_commit (com_total=82, com_tx=41)
B statistics db_con_rollback (roll_total=0, roll_tx=0)
M ***LOG Q02=> wp_halt, WPStop (Workproc21 4256) [dpnttool.c 333]
B Mon Jan 25 15:55:06 2010
B dbmyclu : info : my major identification is 3960472447, minor one 804.
B dbmyclu : info : Time Reference is 1.12.2001 00:00:00h GMT.
B dbmyclu : info : my initial uuid is DF09FC4C150615F194610015C5FDA6C1.
B dbmyclu : info : current optimistic cluster level: 3
B dbmyclu : info : pessimistic reads set to 2.
M Mon Jan 25 15:59:49 2010
M *** WARNING => PfSetGuiTime: gui time too large, discard (gui=-608670296/resptime=0)
Please let me know the preventive steps that we can take to avoid this situation.