How to set up clusters in R server so that each R shells in the Web service calls can access it? RRS feed

  • Question

  • I am publishing R codes as web services in Machine learning server using the mrsdeploy package. I need to implement parallel processing.If I am setting up the clusters within the code, each time the API calls are made 2 sec is lost to setup the clusters. This cannot be done because the entire code is supposed to run within 1 sec(Which is the reason why I am trying out parallel processing).Is there any way to set up the clusters separately and make it available across all R sessions?

    I used the SOAR  package and made the cluster object available across all R sessions a

    But this caused the below error

    Error Error in summary.connection(connection) : invalid connection

    I tried  deregistering the cluster

    But that didn’t fix the issue

    Friday, July 13, 2018 4:12 PM