Oracle and Cloud Wars

I just finished an interview with a friend for a blog post where he was surprised how differently I was viewed in the Oracle community, (a multi-platform DBA who wasn’t very loyal to Oracle like she should be) vs. the Microsoft community, (that odd Oracle girl.)

The truth is, I’ve always enjoyed database technology and few know that my first database (MSSQL 6.5) as well as my first book was in Microsoft SQL Server (Pro SQL Server 2012 Practices).  I’m not platform centric at all, work mostly in Oracle, SQL Server/Azure SQL and MySQL, but I’m just CUSTOMER centric.

Azure and Oracle

I’ve always been here for what the customer needs, (not what they want, which are often two very different things.)  While I was at Oracle, the Oracle Public Cloud was all about Oracle workloads and there was very little to focus me outside of this in customer engagements.  Once I came to Microsoft, the Azure cloud was a monstrous landscape-  an enterprise level cloud that many wanted to bring all their data estate to, but few understood Oracle to know how to migrate it properly.  Many times, folks can’t even navigate the conversation even though the Azure IaaS is a great home for Oracle workloads.  For this reason and this reason alone, I hung up my hopes to update much of my Microsoft knowledge and focus on the areas that helped Oracle succeed in the space.

Marketing, Marketing

Much of the marketing, as we all know marketing will be, is product focused-

Learn how to save on licensing costs and migrate your Oracle database to PostgreSQL/Aurora!

Leverage the power of Azure/AWS/GCP cloud with XXXX and migrate off your Exadata!

Attend this free webinar and learn how to remove Oracle blockers and migrate to XXXXX today!

Now there’s a number of problems with the statements above.

1. They’re forgetting that an Oracle database is often one of the highest revenue creating databases in a customer’s data estate and the loss of that revenue and time to refactor can be unacceptable to the business.
2. Refactoring the database and application doesn’t mean you’ve succeeded. I see many refactoring projects forget that it also has to perform.  If the code and design lends to a specific platform, then you may need to do a lot more work than just translating it.
3. The idea that migrating off of Oracle will lock them into a cloud rarely, if ever works, give up the myth, marketing-
   1. We migrate workloads off of AWS/OCI/CGP all the time. Azure PostgreSQL/ GCP PostgreSQL/AWS Aurora, it’s all open source and can be migrated- easy.
   2. Hate to tell you this, but no matter where you port the workload to, it’s still a form of SQL and yes, we can migrate that elsewhere, too. If you can refactor it once, you can do it again and far easier after you’ve done the heavy lifting off of Oracle.
4. Oracle blockers? The blockers are often boiled down to what marketing doesn’t feel are important, but the customer finds quite overwhelming.  The databases are huge and the idea that you can consume this into another offering, PaaS or otherwise, is very attractive to product groups from any of the cloud vendors.  Oracle is more than 30% of the relational data out there.  Even my third-party partners are quite accustomed to being brought in to lift and shift a customer out from under a competing cloud that is attempting to refactor.

Customer Focus

What really drives refactoring?  The customer.  If the customer wants to refactor from Oracle, then I will support them and help them in anyway I can.  At Azure, I’ve been identified as one of those that can help the customer migrate Oracle in a way that:

1. Lifts and shifts the workload, not the previous onprem infrastructure, which saves considerably on licensing 80%+ of the time and making it easy to know what resources are required to run Oracle in Azure.
2. Promotes Oracle supported architecture in a third-party cloud.
3. Architects Oracle in a compatible way with the HA built into Azure.
4. Has the know-how to build out the support tools such as Oracle Cloud Control, scripting and automation to ease the management and “simulate” a PaaS solution in IaaS.
5. What pitfalls to avoid during a migration.
6. How to get long-term satisfaction in Azure with Oracle.
7. How to evolve the data estate as time goes on to use more Azure services to do more with that data. 
8. Also, refactoring smaller databases or even just schemas is much easier once you’re already in the cloud.

This is a customer first approach, not a marketing one.  Quotas?  Product focus?  Pffftt…that just comes naturally when you give your customers what they need.