Over the past year, I’ve hosted a number of cloud industry panels on the concerns, challenges, and (hopefully) benefits of moving applications into the cloud. I always start the panel discussion by asking for opinions on the No. 1 challenge facing an enterprise when thinking about moving to the cloud, but then each panel session takes off in a different direction based on the panel members, the audiences, and what concerns or news stories are most prevalent at the time.
For example, my first panel discussion of the year focused on auditing and compliance issues with cloud storage (a good story I’ll save for a future column), some panels discussed the real-life relevance of cloud bursting production applications into the cloud, others debated the pros and cons of internal vs. external clouds, and some focused on new(er) cloud models such as community clouds for vertical markets.
But the one theme that all the panel members and discussions kept circling around was how to manage this new cloud beast. It is entirely possible that I somehow directed each panel to discuss management challenges – it is my favorite cloud discussion topic, after all – or it’s possible that the abstract nature of running applications in the cloud makes people start to bubble up the conversation a bit and to focus on the ins and outs of controlling all these new moving pieces. Either way, every panel discussion offered up new topics, new ideas, and new ways of solving the daily challenges facing cloud adoption.
As each panel discussed these topics and ideas, we would drill down into the technical challenges of very specific management issues, such as the lack of a common API framework for cloud provider integration or how logically and physically to separate the multi-tier application architecture – presentation, logic, and data – between on- and off-premises cloud deployments (and if that separation even makes sense).
At the same time as dealing with the technical issues, though, many people took the conversation in the other direction, out of the weeds, and to more macro management architecture issues for the cloud. Unsurprisingly there was a fair amount of discussion about how to use the cloud effectively for disaster recovery. On the surface this model seems like a gimme: Only invoke a cloud deployment in the case of last resort. With the exception of dev and test, using the cloud for DR is basically the quintessential value proposition for enterprise IT. This idea clearly shows that they are both embracing the value of cloud computing while also trying to grapple with new integrated cloud-based architectures. There’s no question that a data center model where all of your applications are ready to go at a moment’s notice but aren’t actually running until needed (and thus not consuming a tremendous amount of operating expenses while dormant). In my mind this is one of, if not the, most compelling use cases of the cloud within enterprise IT.
Practically speaking, however, we fall back to the management challenges of such a highly available model. There are a tremendous number of moving pieces that need to be pre-built, deployed, and managed in real time for a cloud DR architecture to be successful. How will I pre-deploy virtual images on the cloud provider’s infrastructure? How do I connect to its infrastructure? How do I keep image templates current to match my internal infrastructure? What about the database back-end and replication? How will my users know when and where to go to access my displaced applications? How will my provider know how to provision my entire DR infrastructure – apps, network, storage, data – when a disaster strikes if my entire primary data center goes off-line? As with all technologies, it’s the minutia that can make or break a project.
To me, these types of discussions are the reasons I love to host and moderate technology panel discussions. They’re always lively and offer an opportunity for new ideas from panel members and the audience, ideas that often start as nuggets and are flushed out into real-world scenarios during the course of the panel. Panel discussions become just that: discussions of problems and solutions rather than a standard one-way presentation. And what better topic to discuss, debate, and debunk than how enterprise IT organizations are using the cloud in the real world. It’s one thing for technologists to talk about everything the cloud can and can’t do – goodness knows we love to both raise and lower the hype flag – but the real stories come from the people who are dealing with these technologies every day. I pitch panel sessions at almost every conference I attend; I just need to make sure people keep talking.
TMCnet publishes expert commentary on various telecommunications, IT, call center, CRM and other technology-related topics. Are you an expert in one of these fields, and interested in having your perspective published on a site that gets several million unique visitors each month? Get in touch.
Edited by Jennifer Russell