Navigating MAS On-Prem with the IBM MAS Infrastructure Calculator

The IBM MAS infrastructure Calculator has been a reliable resource to determine the compute resources needed to host IBM Maximo Application Suite (MAS). Given the MAS components expected to be used, numbers of Users, and how the RHOS environment is provisioned, it provides a reasonable estimate of the compute and memory needed. 

While it is useful for clients and their cloud hosts to estimate the infrastructure needed (and costs) for MAS instances in the cloud, it can also be used for on-prem estimates. Keep in mind that the calculator does not cover all the costs that an on-prem situation will incur. 

Calculator starting point 

Let’s assume a MAS instance that is only going to serve the MAS Manage component to 100 concurrent users using a fresh OpenShift environment, accessing the current underlying database, and supporting 2 non-production instances. (We will use the MAS 9.0 version of the calculator) Entering these values results in the need for 46.5 vCPUs, 190 GiB memory, and 840 GiB of storage space. An “x-small” environment. 

Each Master Node (a server), and there needs to be 3 of these, requires 4 vCPUs, 16 GiB of memory, and 120 GiB of storage. The average cost per physical server that would meet these specs is $1500 each. 

Each Worker Node (a server as well), 2 per Master Node, requires 16 vCPUs, 64 GiB memory, and 400 GiB of storage. 

What does this mean? 

This is a total of 9 servers @ $1500 each for a total hardware cost of $13,500. 

Whether the servers are bare metal or virtual, the CPU, memory, and storage required to support a production instance of MAS Manage with two non-production instances have a cost to operate on prem. 

Massaging the numbers 

Given these baseline compute requirements, it is an easy matter to adjust the number of Users, the MAS components that will be deployed, and even the number of non-production environments. Any changes beyond the base configuration of “just” MAS Manage will cause the compute requirements to increase along with the costs. 

If we step up the 100 users upwards to 200 users, the results do show an expected uptick in the compute needed, but it is not dramatic. 

For a more significant example, adding the deployment of the MAS Monitor component for 5 users increases the environment to a “small,” requiring 179.5 vCPUs, 550 GiB memory, 2805 GiB storage, and 2048 GiB db2 database accommodation. 

Of course, this calculator is going to reflect the “bare minimum,” so adding 30-50% more of what it is suggesting may be prudent for an on-prem installation. Best to start out with a bit too much compute than to have reports of slow response times from the user community when the shiny new solutions see the light of day. 

Next steps? 

For an on-prem installation the hardware is just the beginning of the costs involved. Typical one-time and recurring tasks/costs will be: 

The above table, based upon real-world estimates, along with the results of the calculator, provides a clearer picture of the costs to operate the RHOS/MAS environment on-prem. If the current legacy Maximo is being hosted on-prem, those servers and operating systems will need to be repurposed or decommissioned as they no longer support the MAS environment. While the concept of hosting MAS on-prem is not much different than hosting the legacy versions of Maximo, the costs can be rather different, mainly due to the hosting of RHOS. Costs are further exacerbated if the IT operation and/or the server farm of the business are contracted out, perhaps necessitating new or renegotiated contracts. 

Is having MAS hosted in the Cloud the solution? 

Hosting your MAS solution in the cloud could be the solution to the many on-prem upgrade issues. Given the user community will not experience any difference in solution functionality whether on-prem or in the cloud, the decision comes down to cost. Not just the initial costs to stand up the environment, but the ongoing operations costs. Whether the cost of going to the cloud is less, the same, or higher will depend on the exact context of the organization’s use of MAS now and into the future. 

Cloud providers offer several approaches that can greatly reduce the initial costs of standing up an on-prem environment for MAS, and turning those remaining into monthly costs spread out over the lifetime of the contract. Shared infrastructure is one method where the cost of some physical elements of the environment is shared across clients. While each client’s data and configurations are certainly kept apart from each other, activities such as solution upgrades and other maintenance tasks may impact more than one client at a time. 

Cloud providers will also take care of the availability, backup, and general maintenance of the environment as part of the typical cloud hosting contract, enabling the client to focus on the use of the solution rather than keeping the environment available. Scheduling patches, assisting the client with validation and other configuration management tasks can also be part of the hosting contract, further reducing the burden on local IT staff. 

Take a hard look at the costs of staying on premises, perhaps overestimating a bit. Then engage with a respectable IBM MAS business partner and/or cloud host to review detailed initial and ongoing cost profiles. Only with all these numbers in front of you can an informed decision be made. 

Wrap up 

TRM has been assisting clients across industries for over 30 years in improving their business processes and maximizing the use of their EAM solution. The upgrade process from legacy Maximo to the IBM MAS solution allows for the opportunity to move from an on-prem situation to one that is hosted in the cloud. There are many advantages (not just financial) to the cloud that might not be apparent, so let us help you work through the details and develop a solution that benefits the business now and well into the future.