Cloud Migration Blog
Back in the late 80's, Cisco, IBM, universities and others started working out the complexities of how to support availability to utilize multiple network paths and vendors with the Border Gateway Protocol.
When it comes to migrating server workloads to the cloud, not all applications are created equal. Mission-critical applications and workloads are typically the hardest to virtualize, and therefore they're the ones you need to be the most cautious about when making changes to the way they are run. They’re often the most complex and most critical to a business and questions around stakeholder buy-in, architecture, and ISV support and licensing must be addressed before making a move to the cloud. Unlike secondary applications, which can usually be virtualized in a self-contained manner, your mission-critical applications touch a wider range of people, processes and technologies.
By Richard Scannell, CEO, RiverMeadow Software
One of the things I get asked about a lot is the size of the market for companies like RiverMeadow - if migration is a “one and done” problem and therefore a limited market size. We have a number of different ways of deriving market size for RiverMeadow, all of which point to a market of about $2B annually in this space – not exactly irrelevant. However, we believe the market, even at that size, is limited today simply because of the limitations of the existing tools. Today, migrations are difficult, risky and expensive – so people do them as infrequently as possible. But, what if migrations were simple, risk free and really cheap – what might happen then?
By Shahin Pirooz, CTO, RiverMeadow Software
As IT professionals, a lot has changed in our world over the last 20 years. With each significant shift, some of us have jumped in with both feet while others have tepidly tested the water with a single toe, and some were dragged along kicking and screaming.
We’ve gone from being the only ones who knew how to build and manage servers to becoming consumers of clouds. We went from being sys admins that did everything once and then forgot it to embracing DevOps to automate our world. We’re no longer perceived as the choke point for implementing new technologies; instead, we’ve become the enablers—the most important part of the product development value chain.
Here we are. Yet this is only the beginning.
Not unlike moving from your current home to a new one, migrating from one Cloud to another can be disruptive, costly and challenging. In fact, to achieve your current Cloud configuration, you had to develop processes, ensure proper personal were in place, implemented the correct set of tools, and contracted with a particular Cloud vendor. Moving your workload and assets to another Cloud provider will necessitate revisiting all of these steps, along with their associated costs. The overwhelming challenge of making such changes can leave many organizations feeling painted into a corner with few palatable choices, essentially locked-in to their existing Cloud vendor.
As today’s leading brand-name hypervisors continue to evolve and improve, adding new features and functionality, the cost of this invaluable technology is also rising. Clearly, the big-name commercially-available platforms offer significant value, providing advanced management, processes and tools, along with a high level of support that creates tremendous comfort for users. However, as organizations look to curb IT costs, exploring open-source hypervisors as an alternate model, it is important to be aware of key differences and limitations.
- Features and Functions May not be Available
The features and functions you have come to know and love in your off-the-shelf hypervisor may simply not be available in an open-source platform. Open-source systems tend to be more immature, which may leave you with the prospect of having to write this missing capability yourself. This alone may change the cost/benefit equation.
Mergers, acquisitions, and even divestitures can present critical opportunities to migrate and consolidate (M&C) systems, allowing an organization to realize cost savings across respective entities. Yet, how the newly combined infrastructures will match-up is seldom a major consideration when consummating these types of transactions. M&A activities can present a compelling case for moving workloads to cloud, but several important considerations should be taken into account.
The reasons for moving business applications to the cloud are plentiful. However, along with the pluses come some minuses. Unfortunately, organizations can run into roadblocks during their migration projects. The critical question is why? As with many business-critical projects, the devil is in the details.
A tremendous amount has been written about the Cloud and how it’s going to revolutionize all of IT. This conversation goes back several years, and while the Cloud is a major driver in consumer-oriented business, (think Apple, Photo Sites) the enterprise has been a little more cautious in it's adoption. Many “greenfield” workloads have been set up in Amazon and other public Cloud providers, however wholesale migration to the Cloud has been lagging.
There are several reasons for this, all of which speak to the challenges that any major paradigm shifts can run in to. With that said, the time is now for entire departments and environments to move to the Cloud. The drivers here may be server refresh, equipment coming off lease, M&A or data center consolidation. Net/Net the same things that have been driving infrastructure change in the past. The major issues that will need to be addressed to facilitate this change are the following:
Successful migration of a workload to cloud is by any measure no small accomplishment. Congratulations are clearly in order. To have made it this far, you developed an effective migration strategy, picked the right migration partner, and chose the right migration platform. Now a larger question looms. What comes next? How can you leverage this success? How can you capture the experience of your successful migration to bring additional workloads into cloud in the future?