App Connect

 View Only
Expand all | Collapse all

ibmint optimize server for node manged server

  • 1.  ibmint optimize server for node manged server

    Posted Mon April 25, 2022 06:48 AM
    Hi,

    Is the ibmint optimize server functionality switched off for node managed servers or just not suitable?

    Background info:
    Our current release procedure is that we release a whole integration server and build a new working directory with all the components in it.
    When done we stop the node managed server, replace the directory under node/server working dir and start the node managed server again.
    So in our situation the content of the node managed server is static.

    This setup has a couple of advantages.
    We don't have to worry about the policies that are as of yet all in DefaultPolicies project.
    Mostly multiple applications in the same integration server are deployed together, they are all stopped and started together.
    Removing an application is as simple as removing the artifact from the release pipeline and release the integration server.

    Having quicker (re)starts (either during release or for any other reason) will be helpfull. We can add the ibmint optimize server during the build of the new working directory.

    ------------------------------
    Regards,
    Jan Fluitsma
    ------------------------------


  • 2.  RE: ibmint optimize server for node manged server

    IBM TechXchange Speaker
    Posted Mon April 25, 2022 07:15 AM
    Hi Jan,

    The ibmint optimize server functionality is currently only there for independent integration servers.

    As you know, the model for an independent integration server is to be able to run inside a container, where the orchestration of updates to code and configuration are handled by the container framework (OpenShift, Kubernetes etc.) rather than through an owning process such as an integration node.  From your background information, it sounds like in your personal situation you are choosing to take advantage of the integration node for process management, but NOT for the purpose of updates to code and configuration... hence why you perceive there to be value in expanding the optimization feature to also include node-owned integration servers.  In our experience, this is not a common pattern - users with the world view of "tear-down and restart" when reconfiguring are very much in the container world. It is for this reason we have not (yet!) attempted to further blur these boundaries by trying to take container concepts and apply them to node-owned servers.  I note your interest here, but this is not currently something we see as popular in the market and we're not currently working towards it.

    Cheers,
    Ben

    ------------------------------
    Ben Thompson
    IBM UK
    ------------------------------