Page 1 of 1

CICS TRACE DETAIL

PostPosted: Thu Nov 25, 2010 2:14 pm
by littlechicken
I have this TRACE record from a CICS dump am I right in thinking its RETURN CODE=1F i.e. 28 for the READ (INVREQ)

from the IBM APRM I get this for READ INVREQ
RC=28
Following a READ UPDATE command without
TOKEN, another READ UPDATE without TOKEN
was issued against the same file without an
intervening REWRITE, DELETE, UNLOCK or
| SYNCPOINT command. This condition may in
| some cases be raised despite the fact that the first
| READ UPDATE was not successful, for example
| because it timed out

The TRACE DETAILS line as follows (ABENDAIDFX formatted in ABBREV form) :

Abend-AID --------------------------------------- Abbreviated Trace Listing ------------------------------- Row 000582 o
COMMAND ===>                                                                                                    SCROLL =
                                                                                                                       
To display the Trace Listing in short format select SHORT                                                               
To display the Trace Listing in full format select FULL                                                                 
To specify selection criteria select SELECT                                                                             
                                                                                                                       
       KE         Trace                                                                                                 
Entry  Task Task  Point   Interpretation                                                                               
000582 0075 61121 AP 00E1 EIP   EXIT  READ                  INVREQ                      01F4,0000001C ....,00100602 ....


Where do I find an IBM manual/webpage with all the TRACE details defined for me please. I have searched till im blue in the face.

I have this link wihch is as close as I can get.
http://publib.boulder.ibm.com/infocente ... fhs14y.htm

am I correct in my interpretation?

Re: CICS TRACE DETAIL

PostPosted: Thu Nov 25, 2010 7:12 pm
by enrico-sorichetti
odd that You could not find anything
this is a link to a cics bookshelf
http://publibz.boulder.ibm.com/cgi-bin/ ... s/DFHASI05

scroll down a few lines and You will find a manual with the title
CICS Trace Entries

which will certainly satisfy Your eagerness to learn about ... "Trace Entries"

Re: CICS TRACE DETAIL

PostPosted: Fri Nov 26, 2010 9:50 am
by littlechicken
littlechicken wrote:I have this TRACE record from a CICS dump am I right in thinking its RETURN CODE=1F i.e. 28 for the READ (INVREQ)

from the IBM APRM I get this for READ INVREQ
RC=28
Following a READ UPDATE command without
TOKEN, another READ UPDATE without TOKEN
was issued against the same file without an
intervening REWRITE, DELETE, UNLOCK or
| SYNCPOINT command. This condition may in
| some cases be raised despite the fact that the first
| READ UPDATE was not successful, for example
| because it timed out

The TRACE DETAILS line as follows (ABENDAIDFX formatted in ABBREV form) :

Abend-AID --------------------------------------- Abbreviated Trace Listing ------------------------------- Row 000582 o
COMMAND ===>                                                                                                    SCROLL =
                                                                                                                       
To display the Trace Listing in short format select SHORT                                                               
To display the Trace Listing in full format select FULL                                                                 
To specify selection criteria select SELECT                                                                             
                                                                                                                       
       KE         Trace                                                                                                 
Entry  Task Task  Point   Interpretation                                                                               
000582 0075 61121 AP 00E1 EIP   EXIT  READ                  INVREQ                      01F4,0000001C ....,00100602 ....


Where do I find an IBM manual/webpage with all the TRACE details defined for me please. I have searched till im blue in the face.

I have this link wihch is as close as I can get.
http://publib.boulder.ibm.com/infocente ... fhs14y.htm

am I correct in my interpretation?


Looks like I am right trace fields are like a minefield to read. they have a logic error which I will sort out with them today.

I actually found a manual yesterday on IBM whichi sort of explains trace fields in more detail. many thanks.

Its not eagerness its a necessity to know more about trace fields hahaha

Re: CICS TRACE DETAIL

PostPosted: Fri Nov 26, 2010 12:09 pm
by enrico-sorichetti
reposting the same info posted in the starting topic is just a waste of resources..

since we spend time in helping You, it would show good manners on Your side
to tell not only that You solved Your issue, but also how You solved it(*)
unless You are one of those greedy people who take without giving
( but that' s quite normal on forums nowadays )

(*) was the manual different from the one I pointed You to

Re: CICS TRACE DETAIL

PostPosted: Wed Dec 01, 2010 2:58 pm
by littlechicken
ok I will post the solution I will go get the trace details etc and post later today.

No I am not greedy I also teach junior MF programmers and spend many hours debugging programs/dumps for people.

Re: CICS TRACE DETAIL

PostPosted: Wed Dec 01, 2010 4:18 pm
by littlechicken
Using ABENDAIDFX as we no longer use IBM formatted DUMPS.

I had to use the trace as the debugging footprint in the program contained footprint of the same value for 4+ cics calls reads or writes to the same database :( - oops.

So tracing back from the *EXC* (the abend on line 506) found the trace record with the READ/INVREQ which looked like this.
000499 0012 29929 AP 00E1 EIP   EXIT  READ                  INVREQ                      01F4,0000001C ....,00100602 ....   
000500 0012 29929 AP 1940 APLI  ENTRY GOTO                  247DC588,COBOL,24084010                                         
000501 0012 29929 AP 1941 APLI  EXIT  GOTO/OK               24010290,24010228,00C50000                                     
000502 0012 29929 AP 00E1 EIP   ENTRY ABEND                                             0004,24084010 .. .,09000E0C ....   
000503 0012 29929 AP 0741 ABAB  ENTRY CREATE_ABEND_RECORD   2408401C , 00000010,XXXXXXXX,YYYY,R14TOR13                       
000504 0012 29929 AP 0742 ABAB  EXIT  CREATE_ABEND_RECORD/OK 27554008                                                       
000505 0012 29929 AP 0741 ABAB  ENTRY START_ABEND           27554008,YES,NO                                                 
000506 0012 29929 AP 1942 APLI  *EXC* Abend-Percolate       START_PROGRAM,XXXXXXXX,CEDF,FULLAPI,EXEC,NO,244B4690,24069CC8 , 00

the entry on line 506 shows the abend occuring.
the entry on line 499 shows the INVREQ condition occuring. I then did a FULL option to get the full trace details after making line 499 at the top of the page :)

XXXXXXXX is the program and YYYY is the abend code.

I looked up manual "dfhs600 cics trace entries zos.pdf" which I have renamed from the IBM short name version and assumed that iis an old style format as it has field A etc.

I then went back to the CICS APRM and looKed up READ and INVREQ and matched FIELD-A(1C) = Decimal 28 and found RC=28 in the READ/INVREQ command RESP2 fields.
RC=28
Following a READ UPDATE command without
TOKEN, another READ UPDATE without TOKEN
was issued against the same file without an
intervening REWRITE, DELETE, UNLOCK or
SYNCPOINT command. This condition may in
some cases be raised despite the fact that the first
READ UPDATE was not successful, for example
because it timed out

and then passed it back to the programmer after determining which exact read failed (a choice of two in the end) - also matched this to the correct READ command in the program.

this is the expanded trace detail
Entry  Task  Interpretation                                                                                                   
000499 29929 AP 00E1 EIP EXIT READ INVREQ REQ(01F4) FIELD-A(0000001C ....) FIELD-B(00100602 ....) KE_NUM-0012 TCB-QR /009E5610
             RET-A4FEC366 TIME-20:01:20.0461598303 INTERVAL-**.**********                                                     
                                                                                                                               


I hope this helps a junior programmer.