Oracle is Doubling Down on Cloud, and on You Doubling Your Data
I just finished watching Larry Ellison’s recent Oracle Cloud announcement. While there are many remarks I could make about the event itself, I will just add a +1 to this article on The Register. Instead the take away I want to share from the announcement is that multi-cloud is here to stay. I know, of all the things that Ellison repeated over and over, that was not one of them; but that is what I heard.
Here’s why. The whole premise of Larry’s speech was two fold. First, they were resolved to slash prices to compete with AWS. No surprise there. Second, they are going to run Oracle Database better than anyone else. This really shouldn’t come as any surprise, but the fact that they have to announce it is telling. All of that said, I expect both of these things as the two primal directives required for their survival. In like manner, I expect Azure to be price competitive with AWS and to do SQL Server better than anyone else. Again, the same two basic primitives. If you do those two things extremely well, then you can compete for my other business like ML, app stack, etc.
But guess what? The days of businesses putting all of their eggs in one basket are over. They want to run ML in Google, MSSQL in Azure, Oracle in Oracle, CI/CD in AWS, and production on prem. Not only that, but they want the ability to pivot any of those decisions at any time. Perhaps they want to shift all Oracle sandbox workloads to Postgres RDS. The commonality between all of this is that data will still exist in multiple places and will need to be copied and moved to multiple places. With the plethora of choice, also comes the multiplicity of data friction. Sure, cloud providers will make it easy to get data into their cloud, and even make it easier to grow that data in their cloud (like Oracle’s Autonomous Database); but it will be incumbent upon businesses to adopt a DataOps strategy that ensures their data can be liberated from any cloud at any time of their choosing.