Is there any guidance on installing FME in AWS? The same question (https://community.safe.com/s/question/0D54Q000080hH8aSAE/installing-fme-server-on-aws) was asked 2 years ago but wasn't answered. Kubernetes and FME Cloud are not options I can use unfortunately.
I would like to use ephemeral machines that can scale with demand and be auto-replaced if they become unhealthy. I'd also like to simply replace the servers using newer versions of the AMI when OS patches or FME updates are released (I don't want long running servers that I have to nurse and treat as pets).
I would like separate auto scaling groups for the engines so I can spin up as many as I want when they're needed and don't have to run more powerful (and expensive) servers when their not required. I might potentially have multiple queues for different types of engine (maybe with different operating systems). I'd need to be able to inspect the queue to know when to scale out/in.
I've looked at doing this (using an external database and file share), but I'm struggling to find the configuration required when I don't know the name of the server that FME Server Core is installed on and the use of random ports between the Engine and Core make load balancing configuration more difficult.
Before I go any further, Is it possible to do any of this? I'm able to get this working if I install the engines on the same machine as the core (following your fault tolerant guidance - https://docs.safe.com/fme/html/FME_Server_Documentation/AdminGuide/Planning-Fault-Tolerance.htm) but I'd really like separate engines to provide flexibility and to reduce running costs. I was wondering whether there was any advice on how to do this. I can build the machine images and infrastructure, I can write scripts to enter the correct variable values into the configuration files on startup, I just don't know which values to enter into the various files when I want the engines to connect to core via a load balancer rather than an individual machine (or if this is even possible). Some advice on how I should approach this would be useful.
Thanks
Allan