DataStage job restore

[jdinero]

I was reading through a lot of posts, but didn't find this question exactly answered.

We run an incremental backup of the unix server where datastage is installed nightly. If today I am working on a parallel job in the client and mess it up, how can I go back to the unix backup and restore the job from yesterday?

I know I can export from the client as a backup, but if I am not doing that, is there a way to restore a specific job from the unix side? I think maybe my question is more about where are the actual jobs are STORED on the unix server side? I know the /Projects folder contains hash files and probably the compiled jobs, but ther is no way to identify them by the name i see using the client.

Can anyone help out or tell me where in the documentation I can find this type of information?

Thanks,

Jennifer

[asorrell]

Jennifer,

If you are doing an incremental UNIX backup there isn't an easy answer. You'd have to track the actual UNIX file that holds the source / object code for your individual job and then recover that specific file. Doing that is risky because you can easily wipe out a job, or worse, corrupt or overlay an entire project if you aren't careful.

I would highly recommend using DataStage utilities to back your jobs up. These can be automated via shell scripts and scheduled to run daily or weekly.

Is this a "what is the best methodology" question or "I've just trashed my job and need to get it back asap" question?

[jdinero]

Thank you, it was a best methodology question!