View previous topic :: View next topic
|
Author |
Message |
william.kangweiliang
New User
Joined: 26 Sep 2008 Posts: 22 Location: China
|
|
|
|
Dear all,
some background before question, there are 5 cics region in my side, 1 TOR, 3 AOR, 1 FOR.
When I try to restart 5 cics, I shutdown them by sequence: FOR -> AOR -> TOR. Is this seq correct?
After use cemt p shut, TOR & FOR normally shut down while in AOR, got such error msg and cics job still run in the SDSF but the CICS session cannot be establish already:
in JESMSGLG:
12.36.06 STC05627 +DFHTM1715 C2TS1A01 CICS is being quiesced by userid SPUSER iin transaction CEMT at netname TCPAB201.
12.36.06 STC05627 +DFHDM0102I C2TS1A01 CICS is quiescing.
12.36.06 STC05627 +DFHCESD C2TS1A01 SHUTDOWN ASSIST TRANSACTION CESD STARTINGG. SHUTDOWN IS NORMAL.
12.36.06 STC05627 +DFHTM1782I C2TS1A01 All non-system tasks have been successffully terminated.
12.36.06 STC05627 +DFHZC2305I C2TS1A01 Termination of VTAM sessions beginning
12.36.09 STC05627 +DFHTD0341 C2TS1A01 512
512 Transaction ILS4 associated with the trigger level for Trannsient Data
512 queue IMMQ has not been initiated.
12.40.05 STC05627 @DFHZC2350A C2TS1A01 534
534 04/23/2009 12:40:05 C2TS1A01 CICS Terminal Control shutdown threshold
534 (4 minutes) exceeded. Sessions still active: C2A1FEAA C2A1FEEAB
534 C2A1FEAC C2A1FEAD C2A1FEAE C2A1FEAF C2A1FEA0 C2A1FEA1 C2A1FEEA2
534 C2A1FEA3 ((1) Module name: DFHZSHU)
in MSGUSR
DFHSZ4154 W 04/23/2009 12:40:08 C2TS1A01 FEPI session setup in pool POOL1 to target TARGETA and node C2A1FE0D failed with a reason
code of X'0801000A'. Setup will be retried later.
DFHAC2007 04/23/2009 12:40:09 C2TS1A01 Transaction ILS4 cannot run as CICS shutdown is in progress.
DFHSZ4154 W 04/23/2009 12:41:08 C2TS1A01 FEPI session setup in pool POOL1 to target TARGETA and node C2A1FE0E failed with a reason
code of X'0801000A'. Setup will be retried later.
DFHAC2007 04/23/2009 12:41:09 C2TS1A01 Transaction ILS4 cannot run as CICS shutdown is in progress.
DFHSZ4154 W 04/23/2009 12:42:08 C2TS1A01 FEPI session setup in pool POOL1 to target TARGETA and node C2A1FE0F failed with a reason
code of X'0801000A'. Setup will be retried later.
DFHAC2007 04/23/2009 12:42:09 C2TS1A01 Transaction ILS4 cannot run as CICS shutdown is in progress.
DFHSZ4154 W 04/23/2009 12:43:08 C2TS1A01 FEPI session setup in pool POOL1 to target TARGETA and node C2A1FE01 failed with a reason
code of X'0801000A'. Setup will be retried later.
DFHAC2007 04/23/2009 12:43:09 C2TS1A01 Transaction ILS4 cannot run as CICS shutdown is in progress.
DFHSZ4154 W 04/23/2009 12:44:09 C2TS1A01 FEPI session setup in pool POOL1 to target TARGETA and node C2A1FE02 failed with a reason
code of X'0801000A'. Setup will be retried later.
DFHAC2007 04/23/2009 12:44:10 C2TS1A01 Transaction ILS4 cannot run as CICS shutdown is in progress.
the problem seems to happen due to fepi session always try to startup while the node C2A1FExx is not available so it always retry and never terminate. Anybody could help to advise how to avoid such problem happen when shut down cics?
My reaction for this case is to use command '/C AORname' to cold shutdown the cics. then I use the 'cold' parameter to startup the cics AOR. It seems work ok. Here, need some master to advise what's the impact on the method I use what's the general process when you restart your PRD cics or UAT cics. Thanks so much!!! |
|
Back to top |
|
|
Garry Carroll
Senior Member
Joined: 08 May 2006 Posts: 1193 Location: Dublin, Ireland
|
|
|
|
Initial reaction is that you should shutdown TOR->AOR->FOR. Shutdown the TOR to prevent any further transactions being initiated, allowing in-flight to complete. Then shutdown AOR since it may have in-flight transactions accessing the FOR. This prevents initiation of new tasks either from TOR or internally. Finally, since nothing can now be accessing the FOR, shutdown the FOR.
Start CICS in the order FOR->AOR->TOR for like reasons.
Issuing /C AORname is drastic and should be avoided!
Garry. |
|
Back to top |
|
|
enrico-sorichetti
Superior Member
Joined: 14 Mar 2007 Posts: 10872 Location: italy
|
|
|
|
I was going to give the same advice,
but Garry beat me on it |
|
Back to top |
|
|
william.kangweiliang
New User
Joined: 26 Sep 2008 Posts: 22 Location: China
|
|
|
|
thanks gary, but if the cics got something like dead loop in the CICS procedure in my case. how to handle without issue '/C cicsname' CMD then? Since I cannot pruge the task in the problemed CICS since the session cannot be establish already. |
|
Back to top |
|
|
Garry Carroll
Senior Member
Joined: 08 May 2006 Posts: 1193 Location: Dublin, Ireland
|
|
|
|
William,
A "tight loop" where you can't establsih a session into the CICS region would suggest that no EXEC CICS or EXEC SQL commands are within the loop. Either of these being in the loop should allow a session to be established. If there are no such interrupts, I'd consider reviewing settings such as runaway task setting (ICVR in SIT). This is used to get CICS to purge such runaway tasks.
Garry. |
|
Back to top |
|
|
|