This article was written by Chris Degnan and originally appeared on the Snowflake Blog here: https://www.snowflake.com/blog/dont-settle-for-multi-cloud-aspire-to-cross-cloud/
Organizations are more often running data and applications on multiple clouds, and that’s great. However, multi-cloud isn’t enough. To let loose the true power of data on your business, you must be cross-cloud.
Cross-cloud means data moves easily between multiple public clouds without any additional work. It means never worrying about where your data and applications live or where your business and technical people are located. With cross-cloud, you get a seamless experience across the major public clouds, so you never end up with the cloud-version of data silos. Most importantly, it paves the way for ultimate business impact—securely accessing and sharing governed data on a global scale.
How multi-cloud became today’s business reality
Before I tackle cross-cloud, it’s important to reflect on how we got to multi-cloud. Snowflake’s journey is no different than any other SaaS provider or SaaS consumer. We started out with the intention to develop and deliver on one cloud provider, in one region, with global deployment.
Certain realities kicked in once our solution went live. Customers wanted us to handle latency issues, match data sovereignty requirements, and help them avoid expenses associated with data egress. This need for proximity to data led us to deploy Snowflake in additional regions within that one cloud provider. As fate would have it, our customers needed support across multiple cloud providers.
This journey to multi-cloud was an organic shift from one region, to multiple regions, to multiple clouds. We went, and to continue to go, where the data is.
Every day, we hear the same multi-cloud story from enterprises large and small, global and local. Whether due to regional regulations or acquisitions, many organizations now find themselves with data sets and applications residing in more than one public cloud.
And, in many ways, this situation is proving to be useful. For global enterprises, multi-cloud addresses challenges such as data latency, single-vendor lock-in, and egress costs. Additionally, it ensures system failover, as well as data resiliency and high-availability.
These multi-cloud benefits are compelling. Nonetheless, there are numerous downsides: a multi-cloud strategy creates boundaries, yields cloud data silos, and enterprises often resort to copying and moving data across clouds. The result? Data-mart sprawl.
Cross-cloud delivers more opportunity than the risk it mitigates
Cross-cloud means data is no longer constrained by the cloud it resides in but rather can move easily across public clouds.
To be clear, this is no small feat. Each public cloud provider developed its own innovative and proprietary way of handling data, which makes it difficult to port data from one cloud to another. Cross-cloud solves this challenge by handling data appropriately with each provider, while providing a seamless end-user experience no matter where data and applications live.
Cross-cloud also ensures business continuity in the event of a technical or catastrophic failure. For example, enterprises in regulated industries require high-availability of mission-critical applications. That means data sovereignty and complete data availability, which can only be provided through mission-critical failover and failback. Of course, the best way to create data replication for failover is by using multiple clouds. If one cloud provider goes down, another cloud provider is used. That’s just part of what it means to be cross-cloud.
But all of this is just table stakes. What transforms a business is modern data sharing, and cross-cloud raises that to a global scale. Modern data sharing enables any two or more organizations to seamlessly share data, unlocking untold insights and new business opportunities that were unimaginable just a few years ago.
You can now seamlessly share data within or across business units, within your ecosystem of business partners, or to buy and sell governed data via a data marketplace. Modern data sharing enables live access between data providers and consumers: data doesn’t move, it’s always up to date, and it’s always available. Cross-cloud takes modern data sharing globally, blurring the lines between regions in a cloud and between clouds. With that kind of access to data, any organization will reveal new business opportunities that were previously unthinkable.
How to become cross-cloud
The only way to achieve cross-cloud is at the platform level. Specifically, today’s modern cloud data platforms enable organizations to access, centralize, and analyze data with near-unlimited scale, concurrency, and performance. These platforms allow organizations to unite their siloed data, discover and securely share governed data, and execute diverse analytic workloads.
To be truly cross-cloud, a cloud data platform must also deliver abstraction at the data layer and be cloud agnostic. The user experience in the data platform should be the exact same at all times, regardless of which cloud provider hosts the data or the application.
As a result, data silos created by individual SaaS applications are eliminated, a single source of data is delivered, and modern data sharing is available locally and globally. The cloud data platform becomes the center of gravity for all data, and organizations can leverage the best of each cloud provider to deliver stronger, faster service to customers.
What that means is freedom—freedom to add, move, or change clouds with ease. This choice becomes a business decision rather than a technical decision. For example, consider an organization that acquires another company using the same cloud data platform but residing in a different public cloud. The data platform makes it simple to consolidate all data into one cloud, or the organization can continue to exist as a multi-cloud, cross-cloud entity. Whatever best suits the business.
Like anything of value, there are costs associated with going cross-cloud, especially around data replication and egress charges. Anytime data is replicated, it costs money and adds a layer of complexity. However, to minimize exposure, a cloud data platform should be smart enough to only send data changes after the initial data sync. This method provides a cost-effective way to optimize customer environments and ensure efficiency in replication.
Enterprises need this flexibility with data because the business environment is demanding it, customers and partners are demanding it, and employees are demanding it. This reality is especially true for SaaS providers, whose users expect data to simply be available and applications to always work.
Application development will advance the benefits of cloud data platforms
Today, a lot of SaaS providers are multi-cloud but not cross-cloud. What that means is, anytime a provider operates on more than one cloud, you might as well be buying two or three different products from the same SaaS provider. The experience will be different depending on where the customer accesses the app and which cloud is used to store the customer’s data.
With cross-cloud, application development takes on a new life. SaaS providers can build applications on top of a cloud data platform and, by nature, access and use data from any public cloud. And, when an application is developed on a cross-cloud data platform, it solves at least two issues:
- Data ownership: When SaaS providers want to take action on data, there’s often a problem: they need to take ownership of that data. However, with a cross-cloud data platform, application developers don’t need to take ownership because the application is simply sitting on top of the data platform. All data is already loaded inside the data platform and can be accessed and used without this additional step.
- Easy deployment: Any time a native app is developed with a backend database or an open-source tool, that app needs to be ported from cloud to cloud and from region to region. However, a cloud data platform allows a SaaS application to sit on any cloud and in any region without requiring additional work. Apps can be deployed once, and from a single location, which means you no longer need to reinvent the wheel every time you do a new deployment, enter a new region, or add a new cloud. And all the data created by individual SaaS apps is available in a single repository.
Both of these points equate to the unique value propositions that a cloud data platform brings to next-gen SaaS app development. Simply develop your app and choose where to deploy it. On any cloud, the customer experience will be the exact same, thanks to cross-cloud.
Remove data boundaries with cross-cloud
Whether through circumstance, experience, or reflection, companies now recognize that multi-cloud is a strong business maneuver. However, cross-cloud is the real game changer and the next logical step for any organization looking to achieve a single source of truth across one or more public clouds.
By selecting a cloud data platform with cross-cloud capabilities, organizations no longer need to consider where data exists or how data access will be impacted by switching cloud providers. Decisions can be made based upon differentiating factors between the offerings, not whether or not data will be siloed in the process.
Personally, I think it’s going to be most exciting to watch SaaS providers take advantage of cross-cloud and build unique applications that are completely agnostic to where the app operates and where the data resides. Now that cross-cloud is a reality, the future is full of opportunities for innovation, which is exactly what happens when data has no boundaries.