It’s a day for celebration! Our migration is complete, and our applications are now running in the cloud environment best suited to their needs. The rest of our application inventory, the ones not cut out for the cloud, remain on-premises where they belong.
We made it!
Actually…we’re not done yet. We still have some work to do to make sure our hybrid environment runs smoothly and delivers the business value we expect.
Fortunately, we aren’t the first ones to travel this path. There’s a lot of good guidance out there we can use, informed by the successes and failures of the cloud adopters who have come before us.
Our existing experience in maintaining smooth operations within the data center won’t go to waste, either. We just need to be mindful of differences, where they exist, and account for them as we create our holistic management strategy.
With that, let’s look at some official guidance from the leading cloud vendors and see whether there’s any consensus as to how cloud operations should be approached.
Vendor-Provided Operations Guidance
In the cloud world, operations encompasses “everything else” that needs to be done after construction and migration are complete. These items include ongoing management, maintenance, and monitoring of the environment, among others.
Anyone with administrative experience, whether data center or cloud, knows a lot goes into addressing these items properly, though. And when these areas are skipped over, bad things tend to happen, so it’s important for operations to receive just as much attention as the architecture and migration phases.
To help address this, the major cloud vendors provide general recommendations intended to help you keep the lights on and avoid the pitfalls of operationalizing a cloud environment.
On the plus side, a lot of this material is well written, so you’ll have no shortage of cloud-based bedtime reading to look forward to.
Amazon Web Services
AWS organizes their guidance into the AWS Well Architected Framework
, which contains five domains referred to as “pillars.” These pillars are broken out as follows, and each has their own comprehensive guide:
The areas most relevant to the operations phase are the Operational Excellence
and Cost Optimization
pillars, so those should be your initial focus.
Of the available vendor guidance, the AWS framework is the most mature, so it’s worth a look even if you’re using a different platform.
Not to be outdone, Microsoft has produced a Cloud Adoption Framework
intended to help customers along the Azure adoption journey. Like the AWS Well Architected Framework, it’s also divided into distinct domains, including:
As you might expect, the Manage domain contains most of what we’ll need to account for during the Operations phase. Within it, you’ll find a good amount of structured guidance on the primary tasks your team will need to handle post-migration.
Areas of Consensus
While each vendor integrates the specifics of their respective technologies into their guidance, there’s a good amount of overlap when it comes to operations best practices across cloud platforms. Some important areas of agreement include (among others):
- Make sure the criticality and business value of each workload is understood
- Be sure all members of the team know their roles and responsibilities
- Monitor critical metrics and continuously analyze for opportunities to improve performance, availability, and cost
- Standardize and automate configuration as much as possible to improve operational efficiency and reduce errors
- Make sure your workloads are secured and only authorized parties have access
- Back up applications and data in alignment with their level of criticality to the business
- Learn from your operational failures and commit to continuously improving skills, tools, and processes
A lot of this probably sounds eerily familiar if you’re accustomed to handling operations within the data center, and that’s a good thing. If you already have a grasp of these general concepts, you’re ahead of the game.
You’ll just need to account for the implications of applying these concepts to a different underlying platform, with a potentially different set of tools and processes, and you’ll be in good shape.
We’ve come a long way in this series, through requirements gathering, environment assessment, platform selection, migration, and now to ongoing operations.
Our business stakeholders are satisfied, and our targeted requirements have been met. We now have a well-designed hybrid cloud environment within which each of our applications happily runs. And we’re in a great position to ensure things stay that way!
It’s now time to start all over at the beginning. Not completely, of course, but reevaluating our requirements and refining our cloud architecture is a never-ending responsibility.
And this should come as no surprise. As IT professionals, we know the work is never done.
So, now that Hybrid Architecture 1.0 is complete, here’s to the successful planning and completion of version 1.1. Cheers!