Amazon RDS, Azure, Editorials

Would You Use a Provider DB Engine?

Perhaps one of the odder questions I’ve asked.  For example, would you use a Microsoft-branded tweaked database platform?  Would you use an Amazon-branded, or Google-branded of the same?

The major players seem to see these as the way forward, and I have to admit it has me both intrigued and baffled.  A strange combination for sure.

With Microsoft, it’s Cosmos DB.  It brings some really intriguing options, like choosing from your level of comfort with regard to consistency (ranging from strong to eventual).  It’s distributed, has some really incredible features and is being touted as the database for the future.  There’s a strong pull to IoT support and scaling in a very big way.

With Amazon, it’s Aurora.  Seemingly built specifically to replace MySQL, Aurora is faster, more fault-tolerant and is compatible with MySQL or PostgreSQL, an interesting ability.  Clearly it’s aiming at their perceived shortcomings for the MySQL and PostgreSQL platforms.

But for years, it’s been a very different story.  It’s been about using platforms that supported standards and workflows and scaling and enterprise levels, etc.  Specifically, I’m talking about SQL Server of course, but you’ve seen similar messaging from other engines.  There are other platforms out there of course, but when it comes to Microsoft, etc. it’s been growing these core and known engines.

So my angst (can one have database engine angst?) comes from wondering about how to choose platforms in the not-so-distant future.  And wondering why it was needed to move to a new platform, to get the scale and functionality.

It almost seems like we’re retreating to very closed vendor solutions, were we railed against those very things for quite a while.

We have been hesitant to use these for our own databases (admittedly perhaps because of ignorance) because of the newness, the move process and the unknowns of what these are truly delivering.  And that’s my question to you.

Are you glad vendors are putting together new database platforms and is it something you could see using for solutions in your business or with your clients?  I’m not here to bash them at all, I’m more interested in the necessary adoption curve and testing and re-development and…. and….

What do you see the biggest benefits of these?  Any drawbacks?  Do you like this approach?

I know the information is just starting to flow (in the case of Cosmos DB for sure) so clearly the jury is still out.

What are your thoughts?