IBM Sterling Transformation Extender

Hi,

in his YouTube-Channel Paul Brett has described how to install ITX Launcher with Docker (Using the ITX Launcher with Docker).

But how would this work if I need to run the Launcher in Multi-Process-Mode?

The problem is that in the folder structure of the customer every system is placed a special folder and the maps also.

Isn't it possible to run the Luncher in Multi-Process-Mode with docker?

Do we have to create a single Container for every Interface and use RedHat Open Shift for orchestration of the containers (about 100 containers in this scenario)?

Maybe it would be easier then just to migrate from Windows Server 2016 to RHEL8 and keep Launcher running under Multiprocess-Mode - don't you think so?

Regards Ben

I haven't done multiple containers in Docker as most of my deployments are done in Kubernetes in the last couple of years.  The most popular ones used by ITX customers are AWS EKS and OpenShift.   However the concepts are the same.  
In K8, you can have multiple containers running, each one in a POD.  In order to obtain AutoScaling, additional POD contains a replica of the original container. The key thing for AutoScaling is sharing of folders (such as maps) among the PODs.  These folders are stored in Persistent Volume outside the container.  In multi-process, if you need to store them in separate folders, you can achieve the same thing but they must be under the same root folder.  In K8, the root folder is /data.  Underneath it, there is maps and systems folder.  This is where you place your compiled maps and msl.

In the maps folder, you can separate them by applications, Perhaps in your case, each process will have its own application name.  Make sure you modify your msd to reflect the location of the maps.