Managing In(and Under) A Cloud

by Adam Hansen, COO & CIO, Assure360

How do you manage a cloud? That may sound like a Zen Koan along the lines of “what is the sound of one hand clapping?”, but it is actually a highly relevant and—in some cases—urgent question for many tech providers, as well as their clients and professional partners.

The instruments of cloud management are still in development, and it is clear that they are not necessarily mature. One particular issue surrounds questions of liability and responsibility: who is being held accountable for downtime or poor performance, and how is that accountability being administered? While many of the details are outlined in the service-level agreement (SLA)—the scope and quality of work, expectations and responsibilities—the specifics regarding performance and penalties often lack clarity. Alternatively, those specifics can skew in favor of the provider in question.

But what happens when a problem arises that is not covered by this agreement with a provider? Or, more commonly, when an issue involves a degree of ambiguity or complexity that makes assigning blame for poor performance a challenge? That question becomes particularly thorny when you consider how to manage those details in a blended cloud arrangement. Keep in mind, in that scenario, you may have assets in a data center, on your own servers, and under the control of ancillary service providers (such as a telecommunication provider that is moving data back and forth between the client and the cloud) that fit into the cloud/service ecosystem. How do you gauge and manage performance when there are so many moving pieces? How do you identify where problems are, and how do you determine who is responsible? Aggregating these dynamics can be a challenge, and understanding what they all mean collectively can be even more of a challenge. 

So what can tech leaders do to overcome these challenges? How can they obtain a clearer picture of a blended cloud service model, ensure that the right parties are accountable in the event of a service interruption or slowdown, and ultimately promote better and more consistent service and value?

Monitor Carefully

Tech providers who are outsourcing a segment of their business to a third party need to be vigilant in monitoring the details of that service. If you don’t have a firm grasp of that, you are exposing yourself to unnecessary costs and (potential) consequences. Review billing and documentation thoroughly: you do not want to make the same mistake that too many people make when they simply pay their phone bill without looking closely at the details of what they are being charged for.

Establish Consequences

To the extent that it is possible, make an effort to make mistakes or instances of sub-par service more painful—or at least more consequential. For that to happen, the consequences of such a lapse must be explicitly stated in the SLA. Be specific about what happens in the event of downtime, remembering to clarify exactly what constitutes “downtime” and how accountability for any lapses will be determined.

Resolve Complexities

The sheer complexity of blended cloud arrangements is a challenge in its own right. Clients, third-party vendors and private cloud providers all have to work in harmony. To that end, CIOs should make sure that the SLA—especially the details about performance and problems—resolves (or at least reflects) these complexities. Reconciling differences between different organizations and recalibrating internal SLAs can be tricky. Even if those metrics are consistent, uncertainties remain. Who is going to hold the vendor accountable if they fall short, and how will that be enforced? Also, because organizations are understandably trying to limit their liability to things they can control directly, the ability to recover damages might be restricted.

Identify Proactive Partners

Because the failure to monitor a blended or private cloud and a lack of specificity and rigor in SLAs are both such common issues, it can be helpful to work with a private cloud provider that has demonstrated the ability to not only monitor their own systems, but also to be cognizant and hold the vendor accountable when they fall short.

Work Hard on the Front End

These issues are not going away. To the contrary, complexity is compounding as cloud accessibility is increasing, and these networks and relationships are only going to become progressively more segmented and complicated. With that in mind, it is essential to plan ahead and be proactive. Start with the end result in mind, and be careful not to “set it and forget it”—monitor your agreements on an ongoing basis, hold parties accountable to the extent that it is possible. 

While there is only so much that can be done, a great deal of the heavy lifting with respect to successful cloud management takes place on the front end, negotiating SLAs that are rigorous, specific and sufficient to protect your interests.