Portal | Manuals | References | Downloads | Info | Programs | JCLs | Master the Mainframes
IBM Mainframe Computers Forums Index
 
Register
 
IBM Mainframe Computers Forums Index Mainframe: Search IBM Mainframe Forum: FAQ Memberlist Usergroups Profile Log in to check your private messages Log in
 

 

How to identify which DSN has an abended record?

 
Post new topic   Reply to topic    IBMMAINFRAMES.com Support Forums -> JCL & VSAM
View previous topic :: :: View next topic  
Author Message
dineshjyrm

New User


Joined: 07 May 2015
Posts: 1
Location: USA

PostPosted: Fri May 08, 2015 6:13 pm    Post subject: How to identify which DSN has an abended record?
Reply with quote

Hi,

I have a job with programs executing in the same and one of the programs is abending with an User abend. The record causing the abend has been identified by performing a dump analysis (Looking at current record read and available in memory).

But it took a long time to identify the DSN in which the record was available since the DD name on in which the faulty record was available had 20+ DSNs concatenated together.

Is there a quick way to identify the DSN in which such faulty record/current record is available??

Thanks,
DJ
Back to top
View user's profile Send private message

Garry Carroll

Active Member


Joined: 08 May 2006
Posts: 990
Location: Dublin, Ireland / Edinburgh, Scotland

PostPosted: Fri May 08, 2015 6:51 pm    Post subject:
Reply with quote

If you know the offset, length and content of the 'faulty' field(s), you could run a DFSORT job against each of the separate DSNs in the concatenation....

Garry.
Back to top
View user's profile Send private message
Terry Heinze

JCL Moderator


Joined: 14 Jul 2008
Posts: 1238
Location: Richfield, MN, USA

PostPosted: Fri May 08, 2015 6:56 pm    Post subject:
Reply with quote

Was there an input record count in the program? If so, you could use the record counts of the 20+ files to find the DSN in question. If there is a unique character string or key, you could SRCHFOR it and it would tell you which relative dataset the record was found in.
Back to top
View user's profile Send private message
steve-myers

Active User


Joined: 30 Nov 2013
Posts: 461
Location: The Universe

PostPosted: Fri May 08, 2015 7:20 pm    Post subject: Re: How to identify which DSN has an abended record?
Reply with quote

dineshjyrm wrote:
... it took a long time to identify the DSN in which the record was available since the DD name on in which the faulty record was available had 20+ DSNs concatenated together. ...
There are two somewhat different ways a concatenated data set can be used: PDS concatenation, and sequential concatenation. Most likely, though, you are using sequential concatenation.
  1. Locate the DCB you are using the read the data set. This is fairly easy for an Assembler program, it is harder for other languages.
  2. Find the DCBTIOT data area in the DCB. DCBTIOT is located at offset X'0028' in the DCB. It is a two byte binary data area. The value in DCBTIOT is the offset of the current TIOT DD entry in the TIOT for the data set currently being processed by the DCB. There is one DD entry in the TIOT for each DD statement.
In a SYSUDUMP type dump, the TIOT is partially formatted, like this -
Code:
TIOT: 009BFFE8                                       
         JOB...... XXXXXXX   STEP..... C             
                                                     
                LN-STA    DDNAME    TTR-STC   STB-UCB
         +0018  14010100  PGM=*.DD  9D2A9800  80F48EA0
         +002C  14010102  SYSUDUMP  9D2D9800  80000000
The +0018 or +002C is the offset of the DD entry. From there you go to your JCL to get the data set.
Back to top
View user's profile Send private message
Bill Woodger

DFSORT Moderator


Joined: 09 Mar 2011
Posts: 7241

PostPosted: Fri May 08, 2015 7:22 pm    Post subject: Reply to: How to identify which DSN has an abended record?
Reply with quote

What are you using to format your abend?
Back to top
View user's profile Send private message
Bill Woodger

DFSORT Moderator


Joined: 09 Mar 2011
Posts: 7241

PostPosted: Fri May 08, 2015 8:08 pm    Post subject: Reply to: How to identify which DSN has an abended record?
Reply with quote

One TIOT entry per DD or one per DSN?
Back to top
View user's profile Send private message
steve-myers

Active User


Joined: 30 Nov 2013
Posts: 461
Location: The Universe

PostPosted: Sun May 10, 2015 6:33 pm    Post subject: Re: Reply to: How to identify which DSN has an abended record?
Reply with quote

Bill Woodger wrote:
One TIOT entry per DD or one per DSN?
The TIOT has one DD entry for each DD statement.

TIOT in SYSABEND or SYSUDUMP -
Code:
TIOT: 009BFFE8                                       
         JOB...... BCT188J   STEP..... B             
                                                     
                LN-STA    DDNAME    TTR-STC   STB-UCB
         +0018  14010102  SYSLOUT   9D26D800  80000000
         +002C  14010102  SNAP      9D2A9800  80000000
         +0040  14010102  SYSPRINT  9D2C1800  80000000
         +0054  14010100  SYSLIN    9D2D9800  80F48D90
         +0068  14010100  APDS      9D2F1800  80F475A8
         +007C  14010100            9D309800  80F475A8
DCB -
Code:
015074 0000     0  00000000 00000000  00000000 00000000  *................*
015084 0010    16  00020000 00000001  00004000 00006AC0  *.......... .....*
015094 0020    32  02000001 54000000  002C0020 009CB038  *................*
0150A4 0030    48  92DF105A 00DF1078  00000001 08000660  *................*
0150B4 0040    64  00000000 00006960  00000001 00000001  *................*
0150C4 0050    80  0000007D 00000001                     *........        *
JCL -
Code:
//B       EXEC PGM=LOADER,PARM='MAP'           
//SYSLOUT  DD  SYSOUT=*                       
//SNAP     DD  SYSOUT=*                       
//SYSPRINT DD  SYSOUT=*                       
//SYSLIN   DD  DISP=(OLD,DELETE),DSN=*.A.SYSLIN
//APDS     DD  DISP=(SHR,PASS),DSN=SYS1.MODGEN
//         DD  DISP=(SHR,PASS),DSN=SYS1.MACLIB
The value in DCBTIOT is 002C, so the DD statement is the SNAP DD statement.

In -
Code:
015094 0020    32  02000001 54000000  002C0020 009CB038  *................*
015094 is the storage address, 0020 is the hexadecimal offset of the address in the block of storage in the dump, 32 is the decimal offset. The TIOT was produced by an Assembler SNAP macro rather than an ABEND dump, but the format is the same, and the storage dump was produced by a custom dump program written for another project.
Back to top
View user's profile Send private message
Bill Woodger

DFSORT Moderator


Joined: 09 Mar 2011
Posts: 7241

PostPosted: Sun May 10, 2015 8:58 pm    Post subject: Reply to: How to identify which DSN has an abended record?
Reply with quote

The TIOT will have one entry for each DD statement, but it will also have an entry for each concatenated dataset on a DD statement. In your example, if X'007C' had been located at that position in the DCB, then it would have indicated the second DSN for the DD APDS, which is what the TS/OP wants to know. They know the DD, but want to know which of 20+ DSNs on that DD contain the current (bad) record.

I think Abend-Aid tells you flat out, if you just find the right place, but I don't know if TS/OP is using Abend-Aid.
Back to top
View user's profile Send private message
steve-myers

Active User


Joined: 30 Nov 2013
Posts: 461
Location: The Universe

PostPosted: Mon May 11, 2015 3:51 am    Post subject: Re: Reply to: How to identify which DSN has an abended record?
Reply with quote

Bill Woodger wrote:
... it will also have an entry for each concatenated dataset on a DD statement ...

Well, it's a DD statement, isn't it?

APDS was a poor example, since it represents PDS concatenation rather than sequential concatenation. In PDS concatenation, once the data sets are opened they operate together; a program using BPAM can quickly and easily switch between the data sets; the currently active data set will not be reflected in DCBTIOT, as is true with sequential concatenation. I put in APDS simply to illustrate what a TIOT with concatenation looks like.
Back to top
View user's profile Send private message
PeterHolland

Global Moderator


Joined: 27 Oct 2009
Posts: 2422
Location: Netherlands, Amstelveen

PostPosted: Mon May 11, 2015 2:28 pm    Post subject:
Reply with quote

The TS/OP is talking about locating the current dataset in a concatenation of datasets that contains the faulting record.

So all the talk about PDS concatenation is of nil value.

Taking a SNAP dump of areas will always show the SNAP DCB as active, because that is where the output goes.

Real abends caused by faulty records and going to SYSUDUMP will give quite
different results concerning control blocks, pointers etc.

So all this is a bit an after the fact situation. The TS/OP has to take care that his input is valid or the processing program can handle rotten input.
Back to top
View user's profile Send private message
steve-myers

Active User


Joined: 30 Nov 2013
Posts: 461
Location: The Universe

PostPosted: Mon May 11, 2015 3:09 pm    Post subject:
Reply with quote

PeterHolland wrote:
... So all the talk about PDS concatenation is of nil value. ...
Yes and no. My only goal was to show the method is not universally applicable.
PeterHolland wrote:
... Taking a SNAP dump of areas will always show the SNAP DCB as active, because that is where the output goes. ...
I just needed an open DCB - and SNAP was the only open DCB. Be reasonable. I needed something quick and one off for the exposition. I was not about to write something where I read part way through a concatenated data set and then took my dump.
Back to top
View user's profile Send private message
Bill Woodger

DFSORT Moderator


Joined: 09 Mar 2011
Posts: 7241

PostPosted: Mon May 11, 2015 3:13 pm    Post subject: Reply to: How to identify which DSN has an abended record?
Reply with quote

Ah.

Code:
// DD ....
Back to top
View user's profile Send private message
PeterHolland

Global Moderator


Joined: 27 Oct 2009
Posts: 2422
Location: Netherlands, Amstelveen

PostPosted: Mon May 11, 2015 9:59 pm    Post subject:
Reply with quote

Quote:
Well, it's a DD statement, isn't it?


That was a stupid reaction.

Quote:
yes and no. My only goal was to show the method is not universally applicable.


Ok, so don't come up with a crap story about PDS concatenation. The same nonsense last time with using Idcams to copy a PDS.

And i'm reasonable, but don't come up with SNAP dumps. Give a real live example, where you find the dataset in a concatenation with the culprit record. That will be a tough job.

So i repeat it once more :

Quote:
So all this is a bit an after the fact situation. The TS/OP has to take care that his input is valid or the processing program can handle rotten input.
Back to top
View user's profile Send private message
steve-myers

Active User


Joined: 30 Nov 2013
Posts: 461
Location: The Universe

PostPosted: Tue May 12, 2015 6:48 pm    Post subject:
Reply with quote

TIOT --
Code:
TIOT: 009BFFE8                                       
         JOB...... XXXXXXX   STEP..... B             
                                                     
                LN-STA    DDNAME    TTR-STC   STB-UCB
         +0018  14010100  SYSLIN    9D26D800  80F48E18
         +002C  14010100  INPUT     9D2A9800  80F48E18
         +0040  14010100            9D2C1800  80F48E18
         +0054  14010102  SYSLOUT   9D2D9800  80000000
         +0068  14010102  SNAP      9D2F1800  80000000
         +007C  14010102  SYSPRINT  9D309800  80000000
DCB -
Code:
0150EC 0000     0  00000000 00000000  0094000B 00F3B03C
0150FC 0010    16  002FE5A2 0501EE68  00004000 00006760
01510C 0020    32  02015074 90000000  00404800 009C958C
01511C 0030    48  12CE90B8 00D30D10  0A000001 02090050
01512C 0040    64  30013030 00006820  0001EF10 0001EEC0
01513C 0050    80  00000050 00000000  00000000 00DE16F8
JCL -
Code:
//B       EXEC PGM=LOADER,PARM='MAP'                         
//SYSLIN   DD  DISP=(OLD,DELETE),DSN=*.A.SYSLIN               
//INPUT    DD  DISP=(OLD,PASS),VOL=REF=*.SYSLIN,DSN=*.A.SYSLIN
//         DD  DISP=(OLD,PASS),VOL=REF=*.SYSLIN,DSN=*.A.SYSLIN
//SYSLOUT  DD  SYSOUT=*                                       
//SNAP     DD  SYSOUT=*                                       
//SYSPRINT DD  SYSOUT=*
Happy?
Back to top
View user's profile Send private message
PeterHolland

Global Moderator


Joined: 27 Oct 2009
Posts: 2422
Location: Netherlands, Amstelveen

PostPosted: Tue May 12, 2015 8:42 pm    Post subject:
Reply with quote

36_2_18.gif
Back to top
View user's profile Send private message
View previous topic :: :: View next topic  
Post new topic   Reply to topic    IBMMAINFRAMES.com Support Forums -> JCL & VSAM All times are GMT + 6 Hours
Page 1 of 1

 

Search our Forum:

Similar Topics
Topic Author Forum Replies Posted
No new posts How to update a portion of text in a ... Bill Woodger DFSORT/ICETOOL 25 Wed Nov 09, 2016 9:41 pm
No new posts sort with previous record anatol DFSORT/ICETOOL 9 Thu Oct 06, 2016 2:36 am
No new posts Get Record count in summary record fo... Atul Banke DFSORT/ICETOOL 21 Fri Sep 23, 2016 4:17 pm
No new posts Change date (DD/MM/YY) in 2nd record ... uday kiran DFSORT/ICETOOL 12 Wed Sep 07, 2016 10:57 pm
No new posts Using 'parm' to vary SORTOUT record v... Sysaron DFSORT/ICETOOL 13 Wed Sep 07, 2016 9:24 pm


Facebook
Back to Top
 
Mainframe Wiki | Forum Rules | Bookmarks | Subscriptions | FAQ | Tutorials | Contact Us