There are plenty of controversial topics in technology to write about. Today I want to combine two of them into one piece: the cloud and naming standards. I’ll leave tabs v spaces and language choices for another day.
I like the cloud, and I especially like Azure. I’m not sure that it fits with most applications and organizations and at times I question the value for some parts of the platform, but overall, the ideas are great. I tend to work more in Azure because I’m an MS stack person, but I suspect AWS and the Google Cloud Platform are similar. Many of the cloud platforms, or even the various services from vendors like Rackspace offer lots of opportunity and an amazing set of platform services.
However, as there are more and more options, tracking the various items and services that you are using becomes an issue. This isn’t a new problem, and I’ve faced the same issue in large Fortune companies. Inventory and management of any technology at scale is hard. This brings me to naming standards. Many of you have probably dealt with naming standards in development and perhaps argued about the choices that your team uses.
There are also issues for infrastructure, which can be more complex and problematic in the cloud. Because we don’t have complete control over devices, we are limited to managing our objects through interfaces. At least in my data center I can fall back on labeling or connecting to a physical box to determine what services are running. Or even turning a system off to see who complains, something I’ve had to do in the past.
In the cloud, the world gets confusing quickly because we have many more types of services. Not only items like VMs and databases, but we can have network interfaces, disks, storage accounts, load balancers, and more. In fact, the list in Azure has gotten extremely long. The addition of things like Azure Functions and Azure Logic Apps have me worried that we’ll start to lose control of our infrastructure. The one thing that I appreciate is Resource Groups, which allow me to put groups of items together.
However, even with these tools, I sometimes find myself confused over what’s running. When I ran across John Paul Cook’s post on naming standards, I realized that I needed to get better organized with my resources. Having some sort of naming standard might help me quickly identify items and keep my portal organized. Even with the various icons, sometimes I get them confused because there are so many different possible icons and services. With a prefix and a resource group, maybe I’ll be better able to track which items I need to keep and which I don’t.
As you grow the number of anything, the overhead for organization and management grows. It becomes more important to track what you have and use, especially when you’re being charged by the resource. I expect that many of us will be answering questions and reporting on not only usage and performance, but also cost for some part of our jobs as more applications move to the cloud.
The Voice of the DBA Podcast