Which Cloud is Best for Hosting my Windows Application?

For Windows® ISVs, choosing the best cloud option for delivering their Windows application is mission-critical. Why? Selecting the “wrong” cloud can affect application availability and performance, compromise application and data security, and negatively affect your bottom line.

Unfortunately, there is no “right” or “wrong” answer when choosing a cloud service. There’s a broad range of factors that can drive the decision, including security, compliance, scalability, cost, and uptime commitment. The best way to choose your cloud is to understand your business needs and goals and select the cloud that best addresses them.

An obvious example is regional coverage. For example, if you have customers across the African continent, Google Cloud™ will not be an option for you since as of this writing it has only one African region cloud center serving the Johannesburg, South Africa region.

Regional coverage is critical for an ISV because regional coverage can significantly affect application latency and cloud cost. The closer the region to your customers’ locations, the less latency those customers will experience when using your application—but the more regions in which you choose to run a workload, the higher your costs. Adding to the confusion is that for some services costs can vary between regions. When comparing providers it’s essential that you’ve identified the geographic regions containing the critical mass of your customers so you can understand how each provider will charge you to run your workload in that region.

Another critical consideration for Windows ISVs is how their application will be able to interact with cloud resources, like storage, processing power, networking, databases, and servers. All public cloud providers offer a software development kit (SDK) that the ISV’s development team can use to optimize the application for that cloud. ISVs should consider the options available from each provider. Additionally, ISV admins should evaluate the administration tools available to manage the cloud environment to make sure that their needs are met.

For most Windows ISVs, the deciding factor(s) will be a combination of elements. For example, if an ISV has a high percentage of seasonal users, the provider’s ability to quickly onboard new users and pricing flexibility related to user count will be critical. ISVs with financial or healthcare applications will need a highly secure cloud, compliance with associated regulations, and strong authorization and authentication safeguards.  

Windows ISVs who are currently delivering app access from a data center that they own (aka a private cloud) may decide to retain their current private cloud when signing up for a public cloud service. A priority for these ISVs could be to select the public cloud that, for example, will enable them to improve application performance in problematic geographical regions and provide monitoring tools that allow their admin team to manage their public cloud in much the same way that they manage their private cloud.

Next, we’ll briefly compare four public cloud providers’ service offerings: Amazon Web Services® (AWS®), Microsoft® Azure®, Google Cloud, and Oracle® Cloud Infrastructure.

Amazon Web Services (AWS)

AWS is the oldest public cloud service, with a 31% market share (as of Q4 2023) and an impressive 200+ individual services to select from. AWS doesn’t provide a standard SLA across their entire cloud; instead, AWS provides SLAs for each service. As an example, the SLA for EC2 (Amazon’s platform for hosting VM instances) is 99.99% at a regional level and 99.95% for an instance. In case of an outage, AWS has a tiered system of service credits tied to outage duration, with longer outages getting larger service credits.

From a data center location perspective, as of this writing, AWS has 33 Regions around the world and 105 Availability Zones. An AWS Region is a cluster of data centers in a specific geographical area; Availability Zones are subsets of regions. AWS Local Zones are extensions of a Region that are location-specific, for example, Boston or Los Angeles. If you use multiple Regions or Zones your overall cloud computing bill will be higher; also, if you need to move app data between regions, AWS will charge you data ingress fees.

AWS provides developers with an SDK to enable applications to more easily interact with AWS cloud resources. For administration, AWS offers a web portal and a command line interface (CLI).

AWS uses a pay-as-you-go model for its cloud resources. Pricing is highly complex—so complex that many consulting practices offer services to help project your costs. Amazon offers an online pricing calculator to help you get started.

Microsoft Azure

Azure is the second-largest public cloud, with 24% of the total market as of Q4 2023, including services like Microsoft 365. Like AWS, each Azure service has its own SLA; uptime guarantees are similar to AWS. Azure’s SLA variable is availability zones—specifically, VMs with at least two instances deployed across two availability zones have a 99.9% uptime SLA; when the instances are in the same availability zone the uptime SLA drops to 99.95%. For single instance VMs, the minimum SLA for any hardware configuration and availability zone is 95%. Microsoft provides service credits when SLAs are not met.

As of this writing, Microsoft beats AWS in regional coverage with 60+ regions versus 33 for AWS. Each Azure region has 3 availability zones, each with its own data center, to reduce latency.

Like AWS, Microsoft provides developers with an SDK and admins with a web portal and a command line interface, although all Azure services and resources can be managed from the portal without using the CLI.

Microsoft claims that AWS is considerably more expensive than Azure when running Windows Server® and SQL Server®. Rather than taking that claim strictly on faith, you can use Microsoft’s pricing calculator to estimate your cost for running your workload on Azure.

{{CTAEMBED_IDENTIFIER}}

Google Cloud

Google has focused considerable effort on acquiring sales talent to grow its Google Cloud market share, now the third-largest cloud platform at 11% as of Q4 2023.

Like AWS and Azure, Google Cloud provides SLAs for specific services which vary based on the service tier. Googe has three service tiers, Standard, Premium, and Enterprise; Google charges for the Premium and Enterprise tiers, which run directly on Google’s Premium backbone. For Google Compute Engine™ (aka virtual machines running on Google’s cloud infrastructure), a single instance in the Premium Tier has an uptime SLA of at least 99.99% in a single zone or multiple zones. Like AWS and Microsoft, Google provides service credits when SLAs are not met.

With the exception of the African continent, Google has impressive data center coverage to support Google Cloud, with 40 regions and over 120 availability zones.

Like AWS and Microsoft, Google Cloud provides developers with an SDK and admins with a web portal and a CLI that supports Linux-style commands.

Also like AWS and Microsoft, Google Cloud’s pricing is pay-as-you-go. Google also offers free credits and access to Google products to organizations who want to try out Google Cloud (this in addition to the Standard Tier, which is free to use up to specific levels). Google also provide a pricing calculator for prospects who want to compare the cost of running different workloads in Google Cloud.

Oracle Cloud Infrastructure

Oracle is the last of this group to enter the public cloud market (in 2016), with 2% of the market as of Q4 2023. The company claims that their late entry gave them the opportunity to learn from the experience of the first to market and enabled them to build their cloud to be more efficient and economical than the established players.

Like the cloud service providers covered above, Oracle Cloud Infrastructure (OCI) has SLAs for each of its cloud services; for example, Oracle NoSQL Database Cloud Service offers an availability SLA of 99.995%. However, Oracle goes two better than the cloud service providers covered above with three SLAs—Availability, Manageability, and Performance—for each service. Like the cloud service providers above, OCI offers service credits when SLAs are not met.

Another area where OCI differs from other cloud services is that they offer cloud regions for commercial customers and cloud realms for government customers. For commercial customers, OCI offers 38 cloud regions in 24 countries. OCI realms are collections of cloud regions that are isolated from each other and do not share data. Realms offered include those optimized for the needs of the US government, the UK government, and the Australian government.

OCI offers a variety of SDKs that developers can use to optimize application interaction with OCI. For admins, OCI provides a web portal and a Service Gateway that enables secure access to multiple Oracle Cloud services from a virtual cloud network (VCN) or on-premises network. 

OCI services are priced the same for every region and realm and vary based on the service used (for example, virtual machine service). Like the other providers covered in this post, OCI uses a pay-as-you-go model, where customers pay for actual use. OCI also offers a Universal Credit annual commitment model, where customers purchase a prepaid amount of credits, which is drawn on based on actual use. OCI offers a cost estimator calculator and a cost management tool available to customers from the web portal. Additionally, customers using another public cloud service can request a bill comparison from OCI.

One Less Decision for Windows ISVs to Make

While each service described above seems similar, there are enough variables between services that selecting the service that works best for you and your customers can be a dicey proposition.

But—once you’ve selected your public cloud provider, it’s easy to select the solution that enables you to share your Windows application with customers located anywhere, on any device.

GO-Global works with every cloud, is easy to install and implement, and leverages any cloud services’ existing infrastructure and security and scalability features to deliver the advanced functionality you need with less complexity and lower cost.

Make GO-Global your go-to. To learn more, request a demo here or download a free 30-day trial.

‍