MVSFORUMS.com Forum Index MVSFORUMS.com
A Community of and for MVS Professionals
 
 FAQFAQ   SearchSearch   Quick Manuals   RegisterRegister 
 ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

Internal restart

 
Post new topic   Reply to topic   printer-friendly view    MVSFORUMS.com Forum Index -> IMS
View previous topic :: View next topic  
Author Message
YSMVS
Beginner


Joined: 19 May 2004
Posts: 51
Topics: 25
Location: My House

PostPosted: Wed Mar 23, 2005 1:27 pm    Post subject: Internal restart Reply with quote

Hi,

Because of the Control interval level locking, we are facing some FD contention abends. We do the restart for the jobs from the last checkpoint. We want to do this in the program without manual intervention. Can any one throw some light on how to do the internal restart from the last checkpoint.

thanks ,
YSMVS
_________________
Thanks,
YSMVS
Back to top
View user's profile Send private message
kolusu
Site Admin
Site Admin


Joined: 26 Nov 2002
Posts: 12377
Topics: 75
Location: San Jose

PostPosted: Wed Mar 23, 2005 8:43 pm    Post subject: Reply with quote

YSMVS,

What exactly do you mean internal restart? Do you mean that the program should restart by it self incase it abends?

If that is the scenario, you will have to do the following.

1. create log sweep job to run every 15 minutes and scan for the abend messages and dynamically generate the JCL and re-submit.

In my opinion this will be a herculian task to sweep the log and generate the dynamic jcl.

2. The other option is to generate the JCL in the abending program itself with restart card last and re-submit it. However you are dealing with contentions , so you need to figure out a way to submit the job so that the job will not abend once again with contention.

Hope this helps...

Cheers

Kolusu
_________________
Kolusu
www.linkedin.com/in/kolusu
Back to top
View user's profile Send private message Send e-mail Visit poster's website
YSMVS
Beginner


Joined: 19 May 2004
Posts: 51
Topics: 25
Location: My House

PostPosted: Thu Mar 24, 2005 8:40 am    Post subject: Reply with quote

Kolusu,

Since we are dealing with the contention abend and most of the time lock gets released immeditely. So What I want to do in the program is, whenever the program faces an FD status code, I want to reposition the database and the GSAM file pointer to the last checkpoint position (by an exteneded restart but without passing the parameter in the JCL) and then reissue the same call. If it has the same FD status code for the same record then abend it to avoid infinite loop, if the call is successful then proceed with program.

I need some gudance on how to do an extended restart from Last checkpoint without passing the checkpoint id from the JCL.

Thanks,
YSMVS.
_________________
Thanks,
YSMVS
Back to top
View user's profile Send private message
Bithead
Advanced


Joined: 03 Jan 2003
Posts: 550
Topics: 23
Location: Michigan, USA

PostPosted: Thu Mar 31, 2005 10:52 am    Post subject: Reply with quote

The abend is occurring in IMS so control is never passed back to the program. Look at using LAST when restarting instead of coding the last checkpoint value.
Back to top
View user's profile Send private message
YSMVS
Beginner


Joined: 19 May 2004
Posts: 51
Topics: 25
Location: My House

PostPosted: Fri Apr 01, 2005 6:54 pm    Post subject: Reply with quote

Bithead,

I am very confused on your reply. Can you please tell me more about the solution you are suggesting.

Meanwhile when I saw the Quick reference for the FD status code, it says that the control is returned to the program

Quote:
FD

Explanation: A nonmessage driven BMP reached a deadlock when IMS tried t
get DEDB or MSDB resources (either DEDB UOWs or overflow latches) for th
program. Or, a mixed- mode BMP reached a deadlock on any resource, eithe
Fast Path or full function. IMS eliminates all database updates that the
program has made since the last SYNCcall, CHKP request, or SYMCHKP comma
(or since the program started processing, if the program has not issued
SYNC call or CHKPrequest). All database positioning is lost, unless you
specified the P processing option in the PCB. Messages to a non- express
alternate TP PCB are discarded.

Rather than having an abnormal termination occur, this status code is
returned to the application program that issued the EXEC DLI command.

Programmer Response: Start processing from the last commit point. If you
reach a deadlock again, terminate the program.


Can you please correct me if I am worng.
_________________
Thanks,
YSMVS
Back to top
View user's profile Send private message
Bithead
Advanced


Joined: 03 Jan 2003
Posts: 550
Topics: 23
Location: Michigan, USA

PostPosted: Mon Apr 04, 2005 11:58 am    Post subject: Reply with quote

YSMVS,

Now I see what you are trying to do. The checkpoint can be passed in the I/O area referenced in the XRST call. Set the I/O area to the word LAST followed by spaces.

See the Application Programming: Database Manager guide for more information on the XRST call.
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic   printer-friendly view    MVSFORUMS.com Forum Index -> IMS All times are GMT - 5 Hours
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


MVSFORUMS
Powered by phpBB © 2001, 2005 phpBB Group