UNIX scheduler product

[Chuck Yerkes]

Quoting Ron Leedy (rleedy at ketera.com):
> Does anyone know of a job scheduler with the following specs:
>  
> - A centralized repository for several servers
> - Flow control (success exits kick off releases another job)
> - Alert e-mails for non-success exits
> - History of scheduled jobs and exit statuses 
> - Solaris and Linux support
>  
> Guess I want my days of working with MVS without paying the Big Blue
> ticket

When Open*Vision bought us (fusion systems group), we had
a scheduler.  They actually acquired another one with hope of
merging them into something unified that worked well.

It was FAR from non trivial (on all Sun clients, run THIS,
if that works, run THAT on certain servers for which the client
run succeeded).  (eg. if the equitiea group's clients succeeded,
run the next step on the equities groups servers.  But don't
touch MBS servers if the MBS client runs failed)

Deep and hard logic as well as some heavy coding.  Veritas bought
OpenVision.  I wonder if either of these products still exist.

That said, cfengine could do some of the lesser tasks described.
(basic distributed cron along with some config awareness).