In data- and analytics-driven organizations, there are almost always analytics workloads with major cost implications. These are the workloads that are—or soon will be—critical to the business but either: (a) cost a lot now, perhaps too much; (b) are growing so as to suggest a coming cost challenge; (c) are candidates for migration to the cloud, to a new platform or a new architecture; or (d) are now being planned for large-scale implementation.
Remarkably, most decisions about such workloads—what platform to use, whether to put them on premises or in the cloud, how to architect them, how to tune or manage their operation in production—are made with eyes wide shut when it comes to cost. Particularly in the cloud, this often leads to large, unexpected bills that can be painful or disruptive or result in outright project failure. On premises, the same phenomenon usually shows up as an unexpected performance problem that then demands a costly hardware upgrade.
We often hear such statements as, “We like the direction of vendor X, so all of our workloads are moving to that vendor’s platform.” Or, “we have chosen our cloud vendor and so we are using the vendor’s native tools for everything.” Or simply, “we have a company standard platform, so that is what we use.”
These approaches are fine for the many small-to-moderate workloads that are not major factors in the cost of operating a business. However, some workloads call for a different mindset and a different approach. For these, there are major opportunities to affect total cost at every stage of the project, from planning to production operation.
This workshop is about how to recognize such situations and how to make effective decisions that take cost and other major factors into account.