Today, more than ever before, Integration is another key enabler behind the use of Cloud.
It is likely, if not inevitable, that remote systems, in different environments, will need to be connected. These systems may be in the form of Software as a Service, "On Premise" or even on separate Cloud infrastructures.
Knowing how to integrate these systems and the most reliable and cost efficient manner to do this is critical in building a performant and scaleable Cloud environment.
There are literally hundreds of integration patterns that can be adopted, and an almost infinite number of data formats. These range from the completely custom through to a huge number of defined data standards.
The standards to be adopted will depend on industry sector and maturity levels. Just some of the common integrtion patterns are listed below;
- RESTFul Apis
- Message Queues
- Protocol Buffers
- Enterprise Service Bus
- Pub/Sub
- ...and many many others
Added to this all of the cloud providers provide levels of capability around integration. Knowing which approaches to adopt and what should be avoided can be a complete nightmare.
Often the integration pattern to be adopted will be influenced by the "Data Rates", how quickly data is being sent to the solution, or the number of concurrent users.
To add to confusion, there are many products on the market that claim to provide full integration services, and these go from competant open source products such as Apache Camel and Commercial products like Mule soft.
Talk to experts who understand how Integration can be delivered quickly and cost effectively.