Checkpoint and restart in IMS DB

pradeepunair · Posted: Thu Sep 06, 2007 3:09 pm

Hi all,
I read through the forum and found this issue with less responses.....

Even i was working with Checkpoint restart program in IMS DB. My code checkpoints and restarts correctly.
But it prints out duplicate records after restarting. I dont know in depth Checkpoints in IMS and got info about it through the posts in this forum.

Can Any of the IMS experts take the initiative to research in this issues and find out the solution.

Thanks for the help :)
Pradeep

itssreehere · New User Joined: 10 Jan 2006 Posts: 65 Location: chennai

Are you getting the duplicates in the output file? Are you using GSAM files? Can you provide more details?

dick scherrer · Posted: Thu Sep 06, 2007 8:10 pm

Hello,

Bitneuker · Posted: Thu Sep 06, 2007 11:57 pm

A checpoint stores your latest position in the database access. If you take one every 100 calls and your program abends let's say after your first checkpoint at call 120 and you outputted the 20 records (not under control of checkpoint/restart) restart will get you back to number 100 and your program will add the 20 records to your outputfile. This is the case if your output isn't controlled with checkpoint/restart.

pradeepunair · Posted: Fri Sep 07, 2007 1:29 pm

Hi all

Thnx for the prompt reply....

My code uses GSAM files. The duplicate records are shown in the output once the program restarts.
Well after analysis I have the following problems with:

1) I analysed a similar piece of code(One program doing a checkpoint and restart). I couldn't make out how the restart logic works out...That is how can i identify a restart in the IMS code.
2) Can i create a scenario where I can have my program restarted. Please help with this.How can I forcibly restart the program to let know the errors its generating.

Thanks
Pradeep

Bitneuker · Posted: Fri Sep 07, 2007 5:20 pm

Seems you need to learn about checkpoint/restart. There are several topics about checkpoint/restart like this one Use the search function to find out about the other ones. You may also follow the first link in my signature and search there for checkpoint and restart. It takes too much to explain all about this matter.

pradeepunair · Posted: Fri Sep 07, 2007 5:38 pm

Thanks Bitneuker
I am searching for more documentation. Most of the links that you have given lead to archived or deleted Info. I am searching about the latest info and once I reach to it i will post it back in this forum.

Thanks for the time and help.

dick scherrer · Posted: Fri Sep 07, 2007 9:06 pm

Hello,

You will probably not find much that is very new concerning checkpoint/restart. Proper checklpoint/restart has been around for more than 20 years and the concept has not changed much since the early days.

Also, this should not be a one person task. An entire databsae environment or at least major applications within it should follow the same checkpoint/restart methodology.

pradeepunair · Posted: Fri Sep 07, 2007 10:30 pm

Hello,

Ya very true. There hasn't been much change in the methodology. But i am unaware if with the new versions of the database there has been changes to the way we try to checkpoint and restart...Has it been more simplified or so????

Anyway as you had commented this shouldn't be a one person job..We have ample posts and replies on this in our forum. We need to standardize this info. I will try from my side...

Also The links by bitneuker were really useful.

Thanks and regards
Pradeep

Bitneuker · Posted: Sat Sep 08, 2007 12:12 am

One might learn from my previous posts; it's my intentention to teach people how to find their way in trouble shooting instead of delivering instant made solutions which would just solve a problem without any learning effect. My guess is that a subsequent link like this one after searching via the first link in my signature at least some understanding of checkpoint/restart will be gained. Whenever problems keep arising don't hesitate to post.