It feels to me like we’re walking away from customers who are critical to our Cloud success.
Development has told me they have been forced into supporting a 4-node cluster as the bare minimum for Fusion Data Catalog (FDC).
I believe this is an incorrect decision because many customers have not yet begun containerizing their workloads but would like to catalog their data as a first step in understanding the full scope of the containerization project. These customers, who are not necessarily small customers, would be the perfect customers for IBM to use IBM Storage Data Cataloging, in either a single node cluster to keep cost and complexity low or a Virtual Appliance (VA) to side-step some of the cost and complexity, as an easy, low stress, steppingstone into containers.
Currently the University of Chicago has no containerized workloads but wants to move in that direction. However, they have no true understanding of their data either. Therefore, they, like many others, would like to use FDC to gain this understanding. However, they cannot justify a 4-node cluster as a first step as the cost and fear are too great. Especially when they can use a far lower cost software product to achieve nearly identical results, with current knowledge.
IBM should design a way for customers to ease into containers and FDC is a perfect solution for this because in the beginning there is usually no TLA on this project and therefore, much less pressure and stress on the staff. Also, performance is generally not a huge issue in the beginning either as it is not yet built into any workflows. IBM could use the transferable version of Fusion software in a single node cluster, or VA, to enable customers to rapidly start the modernization project in a very easy manner. Then, as they add containerized workloads to their environment, backup is a natural fit (although backup capability of Fusion should be a part of the FDC solution itself, there should be no charge until another workloads is added) and a cluster could be migrated (grown) into a full Fusion HCI, provided we used the proper node as the single node cluster for FDC. We already know Fusion HCI is a great starting place, we just need to make it easier for customers to begin in this new environment.
To compound the frustrations in this area, adding FDC to all ESS sales has not been presented well enough and now some customers (UofC is one) think they have FDC. However, they only have entitlement to access FDC not the entitlement to install it.
In my opinion, we have an opportunity to catch a much larger portion of cloud market if we aim for the starting point of building the first containerized workload.
Fishermen catch smaller fish and use them as bait for the larger fish. I think IBM is entering an era where this philosophy will be strategically critical, and if squandered, could be devastating to the company given our direction. Customers are already making cloud decisions daily.
I’d like to ask if we can reconsider our strategy for FDC only deployments?
Call me a wild duck or whatever, but this could be a game changing play that sets IBM apart by giving customers a way to begin their entire cloud journey in a single offering that integrates with existing data, provides their first containerized applications, blooms into a full IBM Fusion HCI, all with low stress, progressive learning, and in a natural growth process. Not to mention reduction in fear of the unknown.
IBM Storage Fusion Data Catalog single node cluster (FDC). This would be the cloud foundation right here.
FDC MES to IBM Fusion HCI by inserting the FDC node into the Fusion HCI rack and integrating it into the HCI configuration. Then integrate the Fusion HCI nodes into the OpenShift environment already in place on the original FDC node.
