I have a bunch of Windows Services and Scheduled Tasks running in one server. Since some of these services pools an external resource I leave it disabled in the backup server, to save bandwidth. How to activate it automatically when the main server comes down in the following scenarios:
This would be a classic failover clustering scenario, so that should be the route to go; but not every application supports MSCS, I know.
The only solution I can come up with is a "controller" service running on the backup server which will periodically poll the main one and start everyting if it doesn't respond for a while.
But this looks simpler than it actually is, and I've seen "redundant" softwares (even quite expensive ones) try this and fail miserably.
If you're using Windows Server Enterprise or Datacenter, you can install a cluster, in which you can configure Services to be clustered Resources. If the active node goes down, the services will be started on the inactive node.
Scheduled Tasks do not look like something that can be clustered, from my experience and reading of the help file. There are 3rd-party enterprise-level scheduling and BPA (Business Process Automation) packages that can handle failures and reschedule jobs on other nodes. You can also try to implement some sort of communication within your scheduled tasks that can determine if a node has failed, and the remaining node needs to start running the jobs.
Windows Cluster Services has a number of other requirements, so it might not be the best fit for you. Without a better description of what your desired end-state is, that's all I can give you for now.
