The total cost of moving your Data Warehouse to the Cloud can be broken into 3 separate buckets. Consulting Services, Hosting and Licensing, and post migration Consumption. Since these coss can vary, we will try to break these costs down for you, so you can understand what you can control.
Migration Services
The cost of migrating to the cloud can vary based on the complexity of the on-prem data warehouse, size of the data, which components need to be moved ad if you are moving to the same or different technology stack. If your entire ecosystem: Database, ETL and BI servers, need to be moved out the current data center, the cost will be prohibitively high. Your IT organization will have to build the business case to move to the cloud. Examples include: Aging hardware, Out of support versions, Data pipeline modernization, or because business needs can no longer be met by the current environment Cloud Migrations can be complex and require planning, resources and time.
DataLake Labs provides end-to-end migration services to help you budget, plan and move to the cloud. We start with discussing your requirements, analyzing your existing environment, and providing you with a Work Breakdown Structure and a cost estimate for services and hosting. We can analyze your current query consumption and usage patterns and predict the total cost of running in the cloud.
Hosting and Licensing
The baseline configuration in the cloud is dependent on the current on-premises system. Architecture decisions have to be made if you are simply upgrading to the same technology stack in the cloud or migrating to a different stack. Your current systems CPU, Storage and IO, all have to be considered in picking a server. Whether you chose to BYOL (Bring Your Own Licenses) or provision a new system in the marketplace, cost are likely to vary. Other considerations are connectivity between on-prem data center and the cloud: Bandwidth, Firewalls, and Security. Addtional costs that can incur are around High Avaialability, Disaster Recover, Multi-Region implementation, Multi-Cloud implementation, and BackUp-Restore solution.
Cloud Consumption
The baseline consumption in the cloud is dependent on the current on-premises system consumption. However, you can either pay-up front and get a discount or pay per-month. In many cases pay per-month turns out to be less expensive than upfront payments because once you have established your run-rate for a month, you know what levers to pull to control and manage costs. In some cases you can scale down all servers to 60% or less when not being utilized.
- Determine Base Line Consumption at 100%, Worst Case Scenario
- Test turning off ETL for one weekend. Friday 7pm to Saturday 7pm. Calculate cost savings for 24 per Week downtime
- Determine what processes could be re-engineered to prevent data movement and egress charges
- Right-size system after first year
- Actively retire workloads that are no longer being used
