Building Enterprise integrations in the Cloud with iPaaS brings many benefits, including among others: simplicity, agility and scalability. However, these benefits should not be taxed by having a weak core, not able to properly manage common enterprise requirements, such as error management. I’ve been a bit disappointed with how most iPaaS vendors handle runtime exceptions of integration flows. A typical example of this, is not being able to support dehydration for asynchronous flows (i.e. dehydration is crucial to supporting long-running instances by saving their memory state into a database, until a correlation invocation, a.k.a call-back, wakes it up to continue with the flow). This causes that when an error occurs, recovery has to start from the beginning of the integration flow that failed.
In these situations, we would have to either design an integration to be fully idempotent and stateless across all its partner links (service invocations), which is not always possible. Another way to do it is by manually handling the recovery of errored scenarios, this is to avoid state inconsistency across the previous service invocations in the orchestration, prior to the error… But then if we have to manually handle compensation, what about iPaaS being easier? Read the complete article here.
For regular information on Oracle PaaS become a member in the SOA & BPM Partner Community for registration please visit www.oracle.com/goto/emea/soa (OPN account required) If you need support with your account please contact the Oracle Partner Business Center.