Step 1: Set up change accumulation batch jobs

The first step is to create batch jobs that define CA groups and identify the spheres in each group.

The example below shows the commands necessary to define a CA group and the sphere names associated with that group when CICS® VR is aware of the logical backups created for the spheres.

Figure 1. Change accumulation batch job when CICS VR is aware of logical backups created for the spheres in the group
//JOBCA1   JOB MSGCLASS=X,CLASS=A,MSGLEVEL=(1,1),REGION=0M 
//* CICSVR PROGRAM                                         
//STEP1    EXEC PGM=DWWCA                                  
//* LIBRARY CONTAINING CICSVR PROGRAMS                     
//STEPLIB DD DSN=DWW.SDWWLOAD,DISP=SHR
// DD DSN=DWW.SDWWLENU,DISP=SHR
//DWWMSG   DD SYSOUT=*                                     
//DWWPRINT DD SYSOUT=*                                     
//DWWSORT  DD SYSOUT=*                                     
//* CICSVR RECOVERY CONTROL DATA SET NAMES                 
//DWWCON1  DD DSN=DWW.DWWCON1,DISP=SHR                     
//DWWCON2  DD DSN=DWW.DWWCON2,DISP=SHR                     
//DWWCON3  DD DSN=DWW.DWWCON3,DISP=SHR                     
//SORTWK01 DD UNIT=SYSDA,SPACE=(CYL,(100,25))              
//SORTWK02 DD UNIT=SYSDA,SPACE=(CYL,(100,25))              
//SORTWK03 DD UNIT=SYSDA,SPACE=(CYL,(100,25))              
//DWWIN    DD *                                            
  CA -                                                     
           GROUP(MYCAGROUP1) -                             
           PREFIX(CA)        -                             
           VOLUME(123456)
           UNIT(3390)
  SPHERE -                                                 
           NAME(PROD.PAYROLL1)                             
  SPHERE -                                                 
           NAME(PROD.PAYROLL2)                             
                                                           
The CA command:
The SPHERE command:

Choose VSAM spheres that have the same backup requirements when you create your CA group. IBM® recommends that you backup all the spheres in a CA group at the same time and then run the CA batch job.

When this job is run, CA calls DFSORT, or an equivalent product with equivalent support, to sort the forward recovery log stream records for PROD.PAYROLL1 and PROD.PAYROLL2. See Diagnosing change accumulation and DFSORT problems for information on eliminating common errors. When the records are sorted, CA consolidates the records by applying the log records for PROD.PAYROLL1 and PROD.PAYROLL2 to a CICS VR dynamically created CA data set.

Every time the CA job is run, the CA data set is updated with the information from the next forward recovery log range.

You do not need to create or keep track of the CA output data set. CICS VR dynamically creates and registers a unique CA output data set using the naming convention, prefix.CAOUTPUT.Dyyyyddd.Thhmmsst, where:
prefix
Is the CICS VR data set name, CICSVR_DSNAME_PREFIX, defined in the IGDSMSxx PARMLIB member or the prefix specified in the CA command. If both are present and different, the prefix in the CA command is used.
CAOUTPUT
Is a constant to mark this as a CA output data set.
D
Is the abbreviation for the date.
yyyy
Is the year.
ddd
Is the day of the year, in Julian format.
T
Is the abbreviation for the time.
hh
Is the hour of the day, based on a 24-hour clock.
mm
Is the number of minutes.
ss
Is the number of seconds.
t
Is the tenths of a second.

When the CA batch job is run again, CA obtains the latest information about the backups. If a new backup has been taken, CA determines that the information in the CA data set for PROD.PAYROLL1 is no longer valid and does not use it. Instead, CA collects the log records from the forward recovery log for PROD.PAYROLL1 and the log records from the CA data set and the forward recovery log for PROD.PAYROLL2 and calls DFSORT to sort the records. When the records are sorted, CA consolidates the records by applying the log records for PROD.PAYROLL1 and PROD.PAYROLL2 to the CA data set.

It is important that you run the CA batch job immediately after a backup is taken for any of the spheres in the MYCAGROUP1 group so the CA information for that sphere is invalidated and is not used by CICS VR if a recovery is requested. Run the CA batch job for MYCAGROUP1 over and over again so that the CA data set is kept up-to-date. CA significantly speeds up forward recovery processing by consolidating the log stream records before a recovery is needed.

For information on diagnosing DFSORT problems see Diagnosing change accumulation and DFSORT problems.

The example below shows the commands necessary to define a CA group and the sphere names associated with that group when CICS VR is not aware of the logical backups created for the spheres in the group.

Figure 2. Change accumulation batch job when CICS VR is not aware of logical backups created for the spheres in the group
//JOBCA2   JOB MSGCLASS=X,CLASS=A,MSGLEVEL=(1,1),REGION=0M 
//* CICSVR PROGRAM                                         
//STEP1 EXEC PGM=DWWCA                                     
//* LIBRARY CONTAINING CICSVR PROGRAMS                     
//STEPLIB  DD DSN=DWW.SDWWLOAD,DISP=SHR                    
// DD DSN=DWW.SDWWLENU,DISP=SHR
//DWWMSG   DD SYSOUT=*                                     
//DWWPRINT DD SYSOUT=*                                     
//DWWSORT  DD SYSOUT=*                                     
//* CICSVR RECOVERY CONTROL DATA SET NAMES                 
//DWWCON1  DD DSN=DWW.DWWCON1,DISP=SHR                     
//DWWCON2  DD DSN=DWW.DWWCON2,DISP=SHR                     
//DWWCON3  DD DSN=DWW.DWWCON3,DISP=SHR                     
//SORTWK01 DD UNIT=SYSDA,SPACE=(CYL,(100,25))              
//SORTWK02 DD UNIT=SYSDA,SPACE=(CYL,(100,25))              
//SORTWK03 DD UNIT=SYSDA,SPACE=(CYL,(100,25))              
//DWWIN    DD *
 CA -                                    
         GROUP(MYCAGROUP2) -             
         VOLUME(123456) -                
         UNIT(3490)                      
 SPHERE -                                
         NAME(PROD.PAYROLL1) -             
         BACKUPTIME(dateandtime,GMT)     
 SPHERE -                                
         NAME(PROD.PAYROLL2) -             
         BACKUPTIME(dateandtime,GMT)     
 SPHERE -                                
         NAME(PROD.PAYROLL3) -             
         BACKUPTIME(dateandtime,GMT)     
 SPHERE -                                
         NAME(PROD.PAYROLL4) -             
         BACKUPTIME(dateandtime,GMT)     
                                   

The CA command defines a CA group name, MYCAGROUP2, and the SPHERE command identifies the spheres in that group, PROD.PAYROLL1, PROD.PAYROLL2, PROD.PAYROLL3, and PROD.PAYROLL4. You must specify the date and time of the most recent backup for the sphere in the BACKUPTIME field. CA only collects records for the VSAM sphere that occur after the backup was made. You can run this job multiple times for the VSAM spheres until a new backup is made. When a new backup is made for a VSAM sphere, you must update the BACKUPTIME field and run this job again. When you run this job, CA calls DFSORT to sort the forward recovery log stream records for PROD.PAYROLL1, PROD.PAYROLL2, PROD.PAYROLL3, and PROD.PAYROLL4. When the records are sorted, CA records are written to a CICS VR dynamically created CA data set on tape volume 123456.


Information Information

Feedback


Timestamp icon Last updated: Friday, 29 November 2013


http://pic.dhe.ibm.com/infocenter/cicsts/v5r1/topic///dwwig/dwwig000126.html