Automation in the cloud: debunking marketing hype

The latest craze in the enterprise job scheduling industry seems to be cloud computing. It can be difficult to determine what a job scheduling product offers to address cloud computing by listening to presentations given by the executives.

• “Support for events is necessary”
• “Linking tasks together”
• “Security”

What is different about job scheduling in the cloud and a traditional enterprise production environment? If you move your data center to Amazon EC2 or another cloud provider, how does the move impact your enterprise jobs?

I was on a conference call with a new Flux customer yesterday discussing how they can leverage Flux more in their cloud environment. It seems pretty simple to me. Servers are static in a traditional data center. Your data center should scale dynamically when moved to the cloud. Today you’re running 5 servers to handle your load. Next month you may need to add 2 additional servers to handle the load. Scaling to meet demand should be automated. Configure the cloud server images with the necessary software and then start/stop instances from APIs or a user interface.

How should job scheduling software support cloud environments? Supporting the ability to add and remove servers dynamically is key.