why do we need job scheduling tool (autosys)

[bunny]

Hi every one,

i am new to datastage, we can schedule jobs using datastage director, then why do we use autosys as job scheduler in datastage

[talk2shaanc]

1. Autosys gives u various options, like JOB_ON_ICE, JOB_ON_HOLD
2. Scheduling is pretty simple, if u hav a job that u want to schedule every one hr. Then through Datastage you have to schedule it 24 times, which would create 24 processes(Distinct PID). whereas in autosys you dont have to take so much pain.
3. if u want to run a job on first monday of every month, u just have to set a Calender in autosys, in datastage couldn't think of.
4. if u want to run a job on first business day(a business day for a client may vary) of every month, u just have to set a Calender in autosys, in datastage couldn't think of.

In short I would say, it would give various scheduling options, with less effort. Reusability and maintenance is also a factor. There many others, i have mentioned just few of them.

If you have worked on Autosys extensively, try to implement on Datastage, certain peculiar scheduling logic that you must have come across in your previous proj using Autosys.

[harithay]

Hello shantanu,

thanks for ur help,

can u give one example how will u connect datastage job with autosys

[ray.wurlod]

Irrespective what scheduler you use (AutoSys, SeeBeyond, ControlM, at, cron, to name a few) use the command line interface dsjob to specify what you want DataStage to do.

The dsjob command is documented in Server Job Developer's Guide and in Parallel Job Developer's Guide - use the index to find "Command Line Interface".

[talk2shaanc]

you will have to write a Wrapper shell scripts, in which you need use Datastage CLI(for details on CLI check the document suggested by Ray).

After creation of the wrapper shell script, you need to just execute that shell script through autosys.