IBM Cloud - Structured Ideas

Welcome to the idea portal for structured ideas (i.e. product feature requests) - A more integrated and automated feedback system to connect your product improvement ideas with IBM product and engineering teams.  Happy submitting!

 

NOTE: All IBM employees must enter Ideas through this Ideas Portal.

Please extend the timeout of the Softlayer loadbalancer

We are using API Connect (APIC) on Bluemix. We have an invoke action on APIC assembly editor and see "500 Internal Server Error" when we invoke a URL which returns about 100MB data . We opened a Bluemix ticket (5377-13438679) to investigate, then it turned out to be caused by the timeout (60 secs) of the Softlayer loadbalancer. The request succeeds if it goes directly to APIC gateway, however, it fails with 500 error if it goes to the the Softlayer loadbalancer. We cannot provide some APIs because of this timeout setting. Please have a look at comment 49 and 50 of RTC Support Ticket 475119 for the details.

  • Guest
  • Aug 17 2017
  • Closed
  • Attach files
  • Admin
    DEEPAK KOTHARI commented
    October 26, 2017 23:49

    Hello,

    It appears that the service you are using is API Connect, which in turn uses some load balancer which is causing an issue. Is that understanding correct?

    Unfortunately, it's unclear which Softlayer Load Balancer is referred to here? and whether it relates to regular Softlayer LB offerings or some embedded load balancer function within infrastructure that API Connect uses? Are you working with anyone in API Connect team? 

    Thanks.

    Deepak

  • Nate Hughes commented
    October 30, 2017 13:36

    Hi Deepak,

    I was working with the client on this issue. This client is on Bluemix Public. Therefore, an adjustment to the SoftLayer loadbalancer timeout would be made by the API Connect operations team and would affect all clients using the environment.

     

    Regards,

    Nate

  • Admin
    DEEPAK KOTHARI commented
    October 30, 2017 18:14

    Hello Nate,

    Thanks. So, it appears that API Connect team is using Softlayer's 'Local Load Balancer shared VIP' service, which is hosted on a shared infrastructure, and hence can't be customized per client without affecting all other clients. Alternatively, the API Connect team may explore using recently-added IBM Cloud Load Balancer service https://console.bluemix.net/docs/infrastructure/loadbalancer-service/getting-started.html#getting-started to see if they run into similar timeout issue. This new IBM Cloud LB uses single-tenant architecture.

    Thanks.

    Deepak   

  • Nate Hughes commented
    November 02, 2017 13:29

    Deepak,

     

    The idea of using an IBM cloud loadbalancer service was brought to the operations team. I'm not personally familiar with that service, but the expert spoken to indicated the service will not work for the client. This problem requires detailed investigation from someone knowledgeable of the client's application architecture.

  • Admin
    DEEPAK KOTHARI commented
    November 02, 2017 18:23

    Hello Nate,

    Ok. The technical folks from the API Connect team need to engage if they aren't already. Since the 'Local Load Balancer shared VIP' service, the service which presumably API Connect team is leveraging, is hosted on a multi-tenant infrastructure, it has limited customization options per tenant. The API Connect team need to investigate other available LB offering options, such as the new IBM Cloud LB service or Citrix VPX LB offering, and use the one that is suitable for their use-case. 

    Adding Savio Rodrigues Offering Manager for API Connect... 

    Thanks.

    Deepak

NOTICE TO EU RESIDENTS: per EU Data Protection Policy, if you wish to remove your personal information from the IBM ideas portal, please login to the ideas portal using your previously registered information then change your email to "anonymous@euprivacy.out" and first name to "anonymous" and last name to "anonymous". This will ensure that IBM will not send any emails to you about all idea submissions