I find it fascinating that when new technologies are invented that are designed to improve efficiency and drive down costs, they end up having the reverse effect – especially the more disruptive ones. Let’s take cloud computing as an example. On the one hand, it provides an enterprise this wonderful ability to offload all of the basal tasks of ordering, installing, and configuring server/storage/network stacks with virtualization and potentially guest OSs on them. This alone is non-trivial and time consuming. For some of the more advanced clouds, you can actually change the size of (virtual) memory, processor, or storage sizes on the fly! What about patching? Hot patching anyone? Automated patches and updates for the core technology such as the guest OSs? These are additional, great benefits found on some cloud providers. And this is just the easy stuff. What happens when a VM becomes a zombie? Who finds, kills, and restarts the zombie? The list goes on …
Now what happens if you realize that one cloud platform was really great for development and scale testing, but when it came to the operational standard that has come to be expected in the enterprise, it can’t be easily replicated on that provider so you decide to build and test on one cloud and deploy on another? Oh and wait … what if you also want to be able to standup just enough to get by on-premises for a last resort, disaster recovery location?
We are seeing customers who currently face these scenarios and they are getting more complex every quarter as cloud companies provide more options, more services, and more choice. Also, this is much more than just having a quick provisioning and automation system. Today’s cloud platform managers all seem to do a pretty good job with that kind of functionality. So how does IT combat this trend that is headed back to siloed management systems and teams and non-conforming governance policies to actually have a way to orchestrate the management stack of very heterogeneous management systems. One company that has focused on this problem is Intigua, which has developed a system that works with many of the top cloud platforms including CloudStack, OpenStack, AWS, and VMware.
Intigua does this via a REST API that enables its policy engine and abstraction layer (SDM) to receive commands from cloud management platforms (vCAC, ServiceMesh, etc.) and provision, configure, and monitor the underlying resources on physical servers, private clouds, and public clouds. This approach allows enterprises to have the ultimate choice in terms of management platform as well as cloud resources and it enables a single method for managing governance policies. The Intigua server REST API also allows integration with the automation scripts such as Puppet and Chef. So when a VM zombies out or dies, the system can sense the problem and remediate automatically where many of the cloud platforms rely on human intervention. As systems require more specific governance policies to ensure compliance, products like Intigua provide a way for centralized management of those policies. Governance policies tied to access rights, protection requirements, and patch levels are just a few examples of some of the basic governance needed across systems. What about who can order resources? What about how much someone can "spend" on resources? How does IT standardize deployment policies across platforms?
As the shift to software defined everything continues to develop and mature, I think we’ll see more of these type of solutions to the problems related to customers wanting choice and best of breed management solutions while still having adequate control over the underlying resources.