IBM Mainframe Forum Index
 
Log In
 
IBM Mainframe Forum Index Mainframe: Search IBM Mainframe Forum: FAQ Register
 

Restart Logic in DB2


IBM Mainframe Forums -> DB2
Post new topic   Reply to topic
View previous topic :: View next topic  
Author Message
madhu.boddu

New User


Joined: 28 Sep 2005
Posts: 5

PostPosted: Tue Dec 27, 2005 3:27 pm
Reply with quote

can any one tell me the restart logic in db2 system.
Back to top
View user's profile Send private message
priyesh.agrawal

Senior Member


Joined: 28 Mar 2005
Posts: 1448
Location: Chicago, IL

PostPosted: Tue Dec 27, 2005 7:37 pm
Reply with quote

Please elaborate more....

Regards,

Priyesh.
Back to top
View user's profile Send private message
madhu.boddu

New User


Joined: 28 Sep 2005
Posts: 5

PostPosted: Thu Dec 29, 2005 7:36 pm
Reply with quote

restart logic in db2, can any body explain in Bref.
Back to top
View user's profile Send private message
bhoompallipraveen

New User


Joined: 07 Dec 2005
Posts: 6

PostPosted: Fri Dec 30, 2005 11:52 am
Reply with quote

What do u mean by restart logic???
Normally we will use check point restart logic in the program while updating/deleting the table or view. It works like this..

for each set of rows say for example 10 u will keep one check pint . If program abends when processing the 10th record you will rool back the changes and next time it has to process from the first.. Suppose it abends at 12th record/row you will roll back to 10th and next time it will process from the 10th record/row.

And plaease note that at each check point you have to issue commit command explicitely.

Hope this will give you an idea

Regds
Praveen
Back to top
View user's profile Send private message
daredevil

New User


Joined: 30 Dec 2005
Posts: 9

PostPosted: Fri Dec 30, 2005 12:51 pm
Reply with quote

Hi Friend,

The restart logic which you gave work ut is difiicult while accessing large volume of records.

suppose if there are around 1000 records, and my program is updating the records when a abend occurs at the 992 record. so updates done till 990 gets commited and last set does not get. the next time when you restart the job, you must do timngs in such a way so that the records starts processing from 991 record.

INstartd of commiting the updates with a cluster of records, you can do a commit each time a record gets updated. since if my job gets abended , i will have all the records before the abended record commited and i can restart the job from the abended record. I am not sure how the performance may vary in both the cases but this is easier. since the DUMP created will hace the abended record details and will be easy to restart the job ( if the job is a critical one and should be restarted in a short time)

Please correct me if i am wrong

Thanks & regards

daredevil
Back to top
View user's profile Send private message
bhoompallipraveen

New User


Joined: 07 Dec 2005
Posts: 6

PostPosted: Fri Dec 30, 2005 4:51 pm
Reply with quote

Hi,

I think your process is quite complicated because you have to commit for each successful record and in case of abend you have to take the record manually to go further.

My case is I will update the table say check-point-restart for each check point commit . next time when I restarted the process I will read this table and get the lattest check point status before proceeding further..
I have done many programs like this!!!!!!!!!!

Hope this will quite eloborate

Thanks
Praveen

[/quote]Live as if you were to die tomorrow. Learn as if you were to live forever
Back to top
View user's profile Send private message
dharmarao.katkuri
Currently Banned

New User


Joined: 11 Dec 2006
Posts: 5
Location: chennai

PostPosted: Mon Apr 07, 2008 11:41 am
Reply with quote

Praveen,

restart logic:
for restart logic wht are the changes we have to made in JCL & COBOL program . can u explain with sample code
Back to top
View user's profile Send private message
dick scherrer

Moderator Emeritus


Joined: 23 Nov 2006
Posts: 19244
Location: Inside the Matrix

PostPosted: Mon Apr 07, 2008 7:11 pm
Reply with quote

Hello,

[quote]for restart logic wht are the changes we have to made in JCL & COBOL program . can u explain with sample codequote]No. This is not a brief subject to be covered by some forum topic.

Backup/cherckpoint/recovery/restart are an enterprise decision/strategy that must be planned and are not considered at the individual analyst or program level.

There are many ways to implement these and all are too "heavy" to try to deal with as a froum topic.

If you have some specific question, someone may have a useful reply, the question you've asked is far too broad. Also, there is much more to a proper restart than a bit of code/jcl change.

I suggest you talk with your seniors on your project as well as the dba(s) who support your environment.
Back to top
View user's profile Send private message
dbzTHEdinosauer

Global Moderator


Joined: 20 Oct 2006
Posts: 6966
Location: porcelain throne

PostPosted: Mon Apr 07, 2008 9:23 pm
Reply with quote

here is an interesting approach to restart.

If you don't want to download it don't, but do not tell me to not attach documents. I can not help it if you work at some paranoid place that has primative security.
Back to top
View user's profile Send private message
View previous topic :: :: View next topic  
Post new topic   Reply to topic View Bookmarks
All times are GMT + 6 Hours
Forum Index -> DB2

 


Similar Topics
Topic Forum Replies
No new posts Finding faulty logic Subscript out of... COBOL Programming 5
This topic is locked: you cannot edit posts or make replies. Need assistance in job scheduling logic. Mainframe Interview Questions 2
No new posts Rexx Logic error while adding seperat... CLIST & REXX 3
No new posts PL/1 Callback address logic in z/OS C... PL/I & Assembler 1
No new posts IMS Restart Failure IMS DB/DC 4
Search our Forums:

Back to Top