Choosing an enterprise cloud platform is a lot like choosing between living in an apartment building or a single-family house. Apartment living can offer conveniences and cost-savings on a month-by-month basis. Your rent pays the landlord to handle all ongoing maintenance and renovation projects — everything from fixing a leaky faucet to installing a new central A/C system. But there are restrictions that prevent you from making customizations. And a fire that breaks out in a single apartment may threaten the safety of the entire building. You have more control and autonomy with a house. You have very similar choices to consider when evaluating cloud computing services.
The first public cloud computing services that went live in the late 1990s were built on a legacy construct called a multi-tenant architecture. Their database systems were originally designed for making airline reservations, tracking customer service requests, and running financial systems. These database systems feature centralized compute, storage, and networking that served all customers. As their numbers of users grew, the multi-tenant architecture made it easy for the services to accommodate the rapid user growth.
All customers are forced to share the same software and infrastructure. That presents three major drawbacks:
Data co-mingling: Your data is in the same database as everyone else, so you rely on software for separation and isolation. This has major implications for government, healthcare, and financial regulations. Further, a security breach to the cloud provider could expose your data along with everyone else co-mingled on the same multi-tenant environment. Excessive maintenance leads to excessive downtime: Multi-tenant architectures rely on large and complex databases that require hardware and software maintenance on a regular basis, resulting in availability issues for customers. Departmental applications in use by a single group, such as the sales or marketing teams, can tolerate weekly downtime after normal business hours or on the weekend. But that’s becoming unacceptable for users who need enterprise applications to be operational as close to 24/7/365 as possible. One customer’s issue is everyone’s issue: Any action that affects the multi-tenant database affects all shared customers. When software or hardware issues are found on a multi-tenant database, it may cause an outage for all customers, and an upgrade of the multi-tenant database upgrades all customers. Your availability and upgrades are tied to all other customers that share your multi-tenancy. Entire organizations do not want to tolerate this shared approach on applications that are critical to their success. They need software and hardware issues isolated and resolved quickly, and upgrades that meet their own schedules.With its inherent data isolation and multiple availability issues, multi-tenancy is a legacy cloud computing architecture that cannot stand the test of time.
The multi-instance cloud architecture is not built on large centralized database software and infrastructure. Instead, it allocates a unique database to each customer. This prevents data co-mingling, simplifies maintenance, and makes delivering upgrades and resolving issues much easier because it can be done on a one-on-one basis. It also provides safeguards against hardware failures and other unexpected outages that a multi-tenant system cannot.
The provider is able to replicate application logic and database for each customer instance between two paired and geographically diverse data centers in each of our eight regions around the world. This can be done in near real-time with each side of the paired data centers fully operational and active. Automation technology can quickly move customer instances between these replicated data center pairs.
It’s important to emphasize that multi-instance is not the same single-tenant, where the cloud provider actually deploys separate hardware and software stacks for each customer. There is some sharing of infrastructure pieces, such as network architecture, load balancers, and common network components. But these are segmented into distinct zones so that the failure of one or more devices does not affect more than a few customers. This enables the creation of redundancy at every layer. For example, at the internet borders, a vendor might have multiple border routers that connect to several tier- one providers on many different private circuits, direct connections, and on different pieces of fiber.
This leads to another important difference between multi-tenant and multi-instance architectures: the approach to disaster recovery. Permanent data loss is a risk inherent to all multi-tenant architectures, and that means external disaster recovery sites are no longer viable options.
True, these are sites that a vendor can fail to if the active side fails. But they are only tested a few times a year and only used if an extreme situation arises. If (when) that happens, they risk failing under load. When that happens, data is lost forever.
That risk virtually disappears in a multi-instance environment. Again, there is not one master file system that services all customers. You can scale out pieces of hardware — stack them on top of each other like LEGO blocks. Each block services no more than a few customers, so one hardware crash cannot affect all the blocks. And because replication is automatic, the secondary side is immediately accessible.
When you partner with a cloud provider that bases its platform on a multi-instance architecture, you’re moving into your own house. Your data is isolated, a fully replicated environment provides extremely high availability, and upgrades on the schedule you set, not the provider. Cloud architecture matters because you’re in control, and better protected when disaster strikes.