It is now well-established that the economics of cloud computing services add massive value to the business, a contention supported both by research and by the number of organisations that either have moved or are now moving production systems into the cloud. Challenges remain of course, and among them is automation.
Cloud certainly changes the economics of IT but it does not necessarily change all the essential nuts and bolts. Systems still need to watched in case they approach their limits, failed hardware needs to be replaced, and applications will still crash or need rebooting. Additionally, of course, the various interfaces between enterprise premises and those of the cloud service provider also need to be configured and managed. There's a host of other tasks too, and many are likely right now to be performed by carefully crafted scripts, each of which needs to be amended when things change, as they do.
All these are essential operations but today's business pressures mean that a faster time to market, fewer human errors, and simplified development and deployment of applications and other resources are driving businesses towards automating as many of these tasks as possible. Research firm IDC reported that the market for datacentre automation software grew by 13.7 percent in 2015, and is expected to be worth $5.7 billion in 2019.
Cloud to the rescue?
So while there can be joy in crafting the perfect complex script, it is not cost-effective for the business to tie up too many of its human resources in this manner, especially when cloud service providers can help by automating many tasks.
Not all cloud providers offer automation as a feature of their cloud services but it is a feature you should be looking for. As an example, the ability to spin up a VM and make a log entry to that effect is one task, while other task bundles or tools may be more complex and nest a range of other tasks, such as truncating a database. Instead of manually connecting to the server, then to the database, checking the size of the database, comparing against the maximum threshold, performing a truncation, and then notifying the user, you might provide essential details such as the names of the server and database and the notification email address, and launch automated tools to do the work.
The cloud provider's tools may not provide a complete solution, of course. While there may be a good deal of overlap, each organisation will have its own processes and expected outcomes, so automated tasks will need to be customised. To help with this, there is likely to be a community around most cloud providers of any size, and help can often be found for practical issues, such as developing your own scripts or customising scripts from other users, as well as tweaks and workarounds.
Given the right automation tools, it is now possible for a single datacentre admin to control and manage hundreds of physical servers (and thousands of VMs), each of them responding in seconds to changes in demand, as a consequence of deploying automated tools to reduce IT's management overhead.
This kind of unified management and automation approach in today's heterogeneous environments means your business will enjoy lower costs, greater flexibility and agility, improved visibility into virtual assets, and fewer 'finger-trouble' errors.
And in the future? Surely the entirely robotic datacentre - a true lights-out environment - cannot be far away.
 IDC: Worldwide Datacenter Automation Software Market Shares, 2015
 IDC: Worldwide Workload Scheduling and Automation Software Forecast Update, 2015-2019