Integration

Get Involved. Join the Conversation.

Topic

    Moon Ray Lo
    OIC Exception/Error Handling & Recoverable InstanceAnswered
    Topic posted February 5, 2019 by Moon Ray LoSilver Medal: 2,000+ Points, last edited March 11, 2019 by Arijit ChakrabortyBronze Trophy: 5,000+ Points, tagged Adapters, Agent, API, Case Management, Cloud, Connection, Database, E-Business Suite, Eloqua, ERP Cloud, Events, Integration, Mapping, NetSuite, Orchestration, PaaS, Primavera, Process, REST, RightNow, Sales Cloud, Salesforce, SAP, Siebel, SOACS, SOAP, Taleo, Tip, Web Services, Workday 
    805 Views, 6 Comments
    Title:
    OIC Exception/Error Handling & Recoverable Instance
    Summary:
    Oracle Integration Cloud's Exception/Error Handling & Recoverable Instance
    Content:

    Hi Team,

     

    I have noticed that NOT ALL instances that faulted can be recovered or resubmitted for reprocessing - based from y experience, I have resubmitted/recovered some (scheduled integration) and unable for others (app-driven integration)

     

    I'd like to know as to which specific conditions would an instance considered as recoverable when fault encountered.

     

    Thanks ahead,

    Moon

    Version:
    19.1.3.0.0 (190129.1200.23460)

    Best Comment

    Moon Ray Lo

    Hi Hemanth,

    I did a POC on this and found out that the documentation is correct, resubmission of recoverable errored instances will start from the beginning and I have attached the POC details in PDF (covered some sensitive data), though I am tagging the User-Managed OIC documentation, instead of the ICS documentation you have provided previously on this thread.

    The mentioned blog above/previously seems to be working on the old version of Integration Cloud which is just standalone ICS (ICS is different from current OIC), this is wayback February 2018 as per date posted - which has much major changes/update from the current version. I tried to replicate this BUT it didn't gave same result om current OIC - the main/parent integration executed the entire 3-minute wait then failed. It seems it worked on that version as the 3-minute-wait acted as Async activity, thus allowing the next activity to execute immediately having ample time to recover, and by recovery on this context, the parent/main instance is still running/in-progress as the user attempted to recover the child - which isn't the case on the current version.

     

    Hopefully these information are enough to close this thread. :)

     

    Thanks,

    Moon

    Comment

     

    • Hemanth Lakkaraju

      Excerpt from the documentation link.

      All faulted instances in asynchronous flows in Oracle Integration Cloud Service are recoverable and can be resubmitted. Synchronous flows cannot be resubmitted.

      • Moon Ray Lo

        Thanks, Hemanth.

         

        With regards to resubmitting errored instances, will it start from the beginning of the integration flow or only start at that errorred part?

         

        Thanks,

        Moon

         

        • Hemanth Lakkaraju

          Excerpt from same document link

          You can manually resubmit failed messages. Resubmitting a failed message starts the integration from the beginning.

          • Moon Ray Lo

            Yes, I have read this also but just want to confirm as someone blogged about this that the orchestration will continue from the last point int failed:

            https://redthunder.blog/2018/02/02/teaching-how-to-recover-errored-instances-with-oracle-integration-cloud/

            "This is very important because that means that any previous successful activity don’t have to be rolled-back or retried, but after a successful recovery, the orchestration will continue from the last point when it failed."

             

            Any idea?

            • Hemanth Lakkaraju

              I have to stand corrected. This looks like a documentation error. Looks like retry will be happening from point of failure. Please raise an SR if you see this is not the behavior on your version.

              • Moon Ray Lo

                Hi Hemanth,

                I did a POC on this and found out that the documentation is correct, resubmission of recoverable errored instances will start from the beginning and I have attached the POC details in PDF (covered some sensitive data), though I am tagging the User-Managed OIC documentation, instead of the ICS documentation you have provided previously on this thread.

                The mentioned blog above/previously seems to be working on the old version of Integration Cloud which is just standalone ICS (ICS is different from current OIC), this is wayback February 2018 as per date posted - which has much major changes/update from the current version. I tried to replicate this BUT it didn't gave same result om current OIC - the main/parent integration executed the entire 3-minute wait then failed. It seems it worked on that version as the 3-minute-wait acted as Async activity, thus allowing the next activity to execute immediately having ample time to recover, and by recovery on this context, the parent/main instance is still running/in-progress as the user attempted to recover the child - which isn't the case on the current version.

                 

                Hopefully these information are enough to close this thread. :)

                 

                Thanks,

                Moon