Many enterprises conduct ongoing IT infrastructure monitoring, but they often lack the critical information needed to support a business decision to move from legacy infrastructure to a public cloud environment, where IT capabilities are delivered externally from an off-premise virtualized environment.
Moving a legacy IT to a public cloud infrastructure offers many benefits: 1) cost savings, 2) improved performance and flexibility, and 3) access to capabilities the enterprise cannot build internally.
However, to realize these benefits, enterprises must change the way they use compute and storage resources.
Running legacy infrastructure is like walking knee-deep in the sea at low tide. As the tide comes in, one has to learn to swim to move effectively and survive in this new environment.
Similarly, in adopting a new public cloud environment, enterprises must learn how to effectively use the cloud before moving there.
This means infrastructure groups have to size compute resources for the cloud correctly in advance, which require a smaller configuration than an on-premise legacy configuration.
Storage has to be configured to take advantage of the various products and tiers available from the cloud providers.
While public cloud decision timing is frequently impacted by hardware refreshes, software license expiration, end of support for operating system security updates, data center lease expirations, etc., a business case should justify the shift as a standalone decision.
To determine whether it makes financial sense to move to the public cloud, enterprises must first establish the current cost of running the infrastructure.
This starts with discovery — including establishing the size and cost of the current legacy environment.
IT may believe cloud technology can speak for itself, however a cost-positive business case is essential in delivering the financial promise and justifying the move from legacy to cloud infrastructure.
IT finance should be involved in understanding legacy infrastructure costs and any contractual obligations that IT stakeholders may not be aware of.
Here are four steps for developing a solid business case:
Enterprises need to establish the granular costs of running legacy infrastructure on a per application or per user level.
The total cost of the IT environment may be known, but perhaps not the unit cost that includes all ancillary and related support services.
This can be accomplished through application and infrastructure discovery, and by linking cloud infrastructure to the cost of each application and resource unit.
The next step is optimizing IT infrastructure by rightsizing the compute and storage resources to identify savings in a commercial cloud. This usually requires a cloud managed services provider or a third-party software package, or both.
Commercial cloud providers usually perform optimization too, which can help IT executives figure out the optimized run rate of legacy infrastructure — thus forming the base case for comparison in the business case.
Enterprises should determine the cost of migrating the legacy infrastructure from where it sits to public cloud infrastructure.
This includes the cost of discovery, optimization, re-configuration and professional services for migration, as well as the training and process change needed.
Enterprises will need to evaluate the cost to build or extend a network core from their current data centers to move infrastructure into the public cloud. There will also be a duplicate operational cost of running legacy infrastructure while standing up the cloud infrastructure during the migration.
Enterprises should take into account that there will be little to no cost recovery from the legacy infrastructure until it is completely turned off, including colocation/data center costs (rack space, power, network, cooling, etc.), and any other contractual obligations the enterprise may have in running the infrastructure.
This transition cost is likely the hardest and least accurate of the numbers in the business case.
Enterprises should consider engaging a cloud managed services provider (MSP) to help with the transition and management of change required — both for the mindset paradigm shift and the organization’s ability to accommodate the change required in running infrastructure in the public cloud.
Timing is critically important in determining whether the business case is cost positive or negative.
Enterprises typically need six to nine months of lead time to perform the steps required to build and analyze a business case, plus another six to 18 months to migrate to public cloud infrastructure.
The business case can change in the space of a few years given the rapid rate of change in cloud pricing and infrastructure cost.
While the business case can be calculated over a one, three or five-year time horizon, a three-year reserved compute instance pricing by commercial cloud providers is usually where the business case makes the most financial sense and where the cost analysis should be done at a minimum.
Inflection points where the enterprise is going through a major transformation are ideal opportunities to move, as these times often reflect the best business case financials.
These may include:
Given the time needed to build a business case around a public cloud shift, this effort should occur at least a year ahead of the inflection point.
In addition, strategic business decisions can significantly change the business case and impact the decision for when to migrate to the cloud. For example, if an enterprise acquires a cloud-centric business, it may be more cost efficient to consolidate and move its legacy infrastructure to the cloud.
The enterprise should conduct the business case analysis, comparing the cost of running infrastructure in the public cloud (and transition) to the base case cost of running legacy infrastructure.
If there is no cost improvement in moving to the cloud, businesses require other compelling considerations to undertake the move.
For example, the public cloud may be able to deliver optimized compute and storage capabilities of the commercial cloud provider.
Enterprises can consider moving only certain functionalities to the cloud or use the public cloud only for services that are available in the public cloud. This includes cloud analytic engines, artificial intelligence and search engines, where enterprises can conduct all computation in the cloud and send the results back to the legacy environment.
Enterprises can also make use of SAP HANA, data warehouses, or cloud-based databases that are delivered via commercial cloud.
This article originally appeared in CIO Dive.