A good number of Maximo Scheduler Optimization customers come to us and ask about:

1. Is there any calculation of the RAM or Hard disc space that is recommended by IBM to provision a MOS server?
2. Is there any recommended OS specification for MOS?
3. Can we run MOS on VPC( Virtual Private cloud) servers?
4. Can we run MOS on Any Cloud provisioned infrastructure?

To be able to answer those question is always good to understand the solution: 


So for a basic setup at the Evaluation phase, our recommendation will be a Docker environment with a minimum of 8GB of Memory RAM, 2 VCPU's, and 100GB of Disk. With that:
- Console will be the small container that will consume fewer resources than the Rest API container, and together 1GB of RAM will be enough to run both together.
- The Execution container is the one where the Demand of Memory can change according to the type of Optimization process (Data, Constraints, and Variables), it is recommended to use as minimum a limit of 8GB of memory RAM and 2 VCPU's with 1 container/pod. End-users will be able to submit 3 parallel Optimization processes at the same time on each Execution Container based on the property MOFEXEC_QUEUEWORKERS (default with 3).
As a container solution, all the parts can be scale up or down depends on the necessity of each customer.
- For the Mongo DB that can be installed with the same docker system running as a container and it will consume more HD space than CPU and Memory, normal we map a dedicated space of 10GB to be able to persist the Optimization Process execution.

Size of the Images: 

*Mongo-seed is the one we use to populate the Maximo Optimization framework with the Model and the Project setting. Just run once and stop.

So after this basic description, we can back to the original question:

1. Is there any calculation of the RAM or Hard disc space that is recommended by IBM to provision a MOS server?
   1. No, we don't have an automatic way to calculate once that needs to be analyzed based on the scenario of Data, Constraints, and Objects. We can affirm with the 2VCPU's 8GB RAM and 100GB HD we are running an Optimization process of thousands of Work Orders (GSLP)
2. Is there any recommended OS specification for MOS?
   1. The base Images of the Container are Linux so we recommend that the Container orchestrator support Linux Base images. Red Hat, Ubuntu, Windows 10, CentOS are OS we already run the solution without problems.
3. Can we run MOS on VPC( Virtual Private cloud) servers?
   1. As long it supports the Linux container that will not be a problem.
4. Can we run MOS on Any Cloud provisioned infrastructure?
   1. Yes, you can run the solution at Any Cloud provider infrastructure. We already deploy at K8s services of IBM Cloud and AWS. That will require some work over the values on the deploy process due to different services for connectivity.

Additional Performance Test results can be found here.


#MaximoScheduler
#Maximo
#AssetandFacilitiesManagement