You are making several confusions. Don't confuse agent mode with scheduler mode. If you setup local jobs and run local jobs you don't use an agent, you use a scheduler mode. An agent is what is used to run jobs submitted from remote schedulers, that's all. Now if you want to run a job on Solaris but start it from Windows and monitor it from Windows you do the following: 1. On Solaris you run 24x7 in the agent mode (user ./agent.sh) to start the agent or ./agent.sh nogui to run it in console mode 2. On Windows run scheduler master.bat and configure an agent profile pointing to your Solaris computer 3. On Windows create a job definition and in job properties select configured profile for Solaris agent as a "Agent/host" to run the job. This way you can run the job from Windows and monitor it from there. Moreover you can have multiple agents and jobs running in many different places and yet link them together using ob dependencies on the scheduler side and again monitor them all in one place, handle errors, run reports, and so on... Now, to make thinks more complicated, if you run a scheduler with "server" mode option enabled it can server at the same time as both a scheduler and an agent, in other words, it can run its own jobs and remote jobs submitted from other schedulers and agents. : I've installed the multi-platform java edition on my Win2K desktop and on : a Unix Solaris 9 server (as a user, not as root). I've started the agent : on the server, and created a job on the server, which runs fine. I've added : the agent to the Windows scheduler, and successfully tested the connection : (the : agent is running on the Unix box). Is there a way to monitor the Unix job : from the Windows scheduler/monitor/log file? Can I create new jobs on the : Windows scheduler to run on the Unix box? How would I do that?
|