Posted: Wed Aug 29, 2007 1:50 am Post subject: Query on CA7 Scheduling
Hi,
Sorry if I am posting this in a wrong forum.
I have the following requirement:
I have two jobs, JOBA and JOBB which are triggered by JOBC. JOBA and JOBB are NDM jobs which send files to external system. This system runs on two LPARS. My requirement is, I have to send the files to two different LPARS on alternate days. For example, if JOBA sends the file to LPAR1 and JOBB sends the file to LPAR2, I want to run JOBA and JOBB on alternate days. I can not schedule JOBA to run on SUN,TUE,THU,SAT and JOBB to run on MON,WED,FRI. This will increase the load on one LPAR.
Could you please suggest any solution?
Thanks,
Ugandhar.
Joined: 26 Nov 2002 Posts: 12370 Topics: 75 Location: San Jose
Posted: Wed Aug 29, 2007 7:26 am Post subject:
yugee,
Code a last step in Job C which can be utility/pgm to determine which job needs to be triggered. Once you determine then you can dynamically demand JOB B OR JOB C using UCC7TRLR. UCC7TRLR is a proc which invokes the PGM SASSTRLR which actually demands the job.
You can find the details of the pgm SASSTRLR in CA-7 Interfaces Guide manual
Hi Kulusu,
I tried to implement this, but as per the standards, triggers insude the jobs are not allowed. So, I am trying for some solution using CA7 scheduling.
Please let me know if there is any option to do this.
Thanks in advance,
Yugee
Joined: 01 Mar 2007 Posts: 475 Topics: 9 Location: Welsh Wales
Posted: Fri Aug 31, 2007 1:52 am Post subject:
I can vaguely recall a similar thing way back.
i.e. a process that was one thing on MON, WED, FRI, and another on TUE & THU
May or may not help you, but a REXX was used to determine which day it was and then copied the appropriate CC member from a PDS to a && file which a later jobstep used for processing options.
Only problem was if the job was severely delayed and ran past midnight. But with CA-7 supplied parms as input to the REXX you can use a logical rather than physical run date. _________________ If it's true that we are here to help others,
then what exactly are the others here for ?
Joined: 26 Nov 2002 Posts: 12370 Topics: 75 Location: San Jose
Posted: Fri Aug 31, 2007 5:35 am Post subject:
Yugee,
I think you can use the SYMETRIC scheduling. The syntax is
Code:
SYMETRIC START: ___ SPAN: ___
For JOB A Start with Jan 1st and with a span of 002. ie
Code:
SYMETRIC START: 001 SPAN: 002
For JOB B Start with Jan 2nd and with a span of 002. ie
Code:
SYMETRIC START: 002 SPAN: 002
Here is the explanation
Code:
SYMETRIC This field is used in combination with the START and SPAN fields to define a symmetric schedule. With this option, schedules are defined as beginning on the date specified with the START field and recurring every nnn days as specified with the SPAN field. The selection of this type of schedule is made by entering an X or Y.
START: This field is used in combination with SPAN and SYMETRIC to define a symmetric schedule. This field is required when the SYMETRIC option is taken. Value must be specified as the Julian day of the year on which the symmetric schedule is to begin. This value should be evaluated yearly before schedule resolution and may need to be changed each year.
If the schedule will be resolved on a July-June basis, the start date must be within the first of the two years in which the schedule overlaps. For January-December schedules, it is simply relative to
January 1 of the year specified on a RESOLV command.
If the calendar against which this SCHID will be resolved does not have the OPTIONS=SCHDYONLY, the START day will be adjusted each year to maintain the job's symmetric schedule from the job's original specified START day.
If the calendar against which the SCHID will be resolved was generated with overall available processing days only (OPTIONS=SCHDYONLY), the start date must specify the first day on which the job would actually run. This requires manually setting (verifying) the START value each year before the
RESOLV.
If a nonprocessing day is specified, the start day is the next processing day found in the calendar. For example, if January 1 is a nonprocessing day and the calendar was generated with OPTIONS=SCHDYONLY, and 1 is specified as the START day,
January 2 is the actual START day (provided it is an available processing day). SPAN is based from January 2 in this example rather than January 1.For other types of calendars, the start date is determined at
RESOLV time based on the ROLL and INDEX options taken.
Size/Type: 3 numeric characters specified as Julian day of the year from 1 to 365
SPAN: This field is used in combination with SYMETRIC and START to define symmetric schedules. This field is required if the SYMETRIC option is taken. When used, specifies the number of days between scheduled processing cycles.
If the calendar against which the SCHID will be resolved was generated with processing days only (OPTIONS=SCHDYONLY), the SPAN value is specified as the number of available processing days between and including the next processing date as opposed to the actual number of days. With this type of calendar, the ROLL and INDEX options serve no practical purpose. For other calendar types, the SPAN value is specified as the number of calendar days between processing cycles and the ROLL and INDEX options can
be used as necessary.
Size/Type: 1 to 3 numeric characters from 1 to 255
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