Sergey Nivens - Fotolia
There's a lot of buzz around multi-cloud and the complexity it presents to enterprises. For many IT teams, that complexity lies largely in the integration work that needs to occur between multiple public clouds but also between on premises and the public cloud.
The reality is: Many multi-cloud environments still include some on-premises components. These include OSes, databases and application servers that both vendors and IT teams have retooled for years to reach out to public cloud. But, though they are starting to improve, public cloud providers haven't traditionally done a great job ensuring their technology can reach back into on-premises systems. Enterprises, as a result, struggle to provide core integrations between on-premises and cloud systems.
Here are three integration approaches admins can take to effectively fold their on-premises resources into multi-cloud environments.
Data integration -- the most straightforward of the three approaches outlined here -- refers to the movement of data from on-premises applications and data stores to their analogs in the public cloud. There are numerous on-premises and cloud-hosted data integration tools, such as Dell Boomi, SnapLogic and MuleSoft, from which enterprises can choose. These tools are sometimes referred to as integration PaaS.
Enterprises need an interface that exists between the data integration server and the on-premises and public cloud-based systems being integrated. These are typically APIs that can interface with ERP systems, databases and other custom or proprietary resources. Data is consumed from the source system, either on premises or in the cloud, and then sent to the data integration server, where it's transformed and then sent to the target system.
Security integration is the ultimate objective for enterprises in multi-cloud environments. IT teams want a single security layer that can work across on-premises and public cloud platforms, which typically means they need a single and centralized directory system, such as Active Directory or other LDAP-compliant options.
These shared directories are the jumping off point for an identity and access management (IAM) strategy that works well across on-premises and cloud-based systems. Once an IAM tool verifies a user's identity, it authenticates that user for most resources across a hybrid or multi-cloud environment, without the need to reauthenticate for each individual database or application. Admins can set up roles for users and themselves, as well as provide and control access to local and cloud-based resources through customizable configurations based on their specific security needs.
Service integration is perhaps the least known on this list. It deals with specific APIs, services or microservices that enable systems to share both behavioral information and data. It's not enough to just stand up these services, though; enterprises need to secure, govern and track them across multi-cloud environments. To do this, use API managers and service repositories to enable the discovery of services and also enforce policies that dictate how those services are used.
Service integration is joined at the hip with server integration and service orchestration tools. These tools help developers bind services together to create more complex processes or applications. For instance, an application that cleans up data once a month would call upon several other services, such as those for data security, verification and validation, to complete that task.