Thread by @QuinnyPig

The first #reinvent launch is in a bit under 4 hours, and with it the start of #reQuinnvent.

Let's get in the mood by reading @gartner_Inc's MQ for Cloud Database Management Systems.

A thread!

If we look at the axis labels, @awscloud's ability to execute is unmatched but their vision trails that of some other companies like @GCPcloud, @OracleCloud, and @IBMcloud.

*spit-takes in Gartner*

Wait, what?!

Okay, @awscloud's strengths make sense to me--but I'll call them out if they start marketing on that "more reliable than other clouds" point too heavily.

You don't market on other people's downtime, for tomorrow it may be yours instead. Or perhaps last week's.

Their cautions are a mixed bag.

First, we blame @awscloud for apparently, I dunno, not shipping the multifunction printer of cloud? What?!

WTF is Gartner asking for here? @awscloud's Database Migration Service (DMS) supports basically any replication target you'd want, including other clouds.

I get that I'm an Enterprise Heretic about this stuff, but "run a database anywhere" isn't exactly common.

Okay, @awscloud's "Day One release quality" is indeed a fair cop. No argument here.

Next big player in alphabetical order is @GCPcloud.
We start with strengths.

Yes, if Google is renowned globally for anything, it's their strong enterprise focus.

Here we go with multi-cloud again.

If Google "has delivered Anthos" then why haven't I seen it in the wild yet?

And this one's fair. "Unless you're a search company or perhaps a mortuary (Hi @killedbygoogle!), we won't directly compete with you, partners."

Now to cautions!

Yes, @GCPSupport is something of a challenge.

Yeah, "how do we leave Cloud Spanner" is one of those questions I like to ask ahead of time. It's great, but the exodus strategy is largely "redesign your software architecture" in many cases.

Wait, didn't you just give these folks better "vision" points than @awscloud? Did the author of this section check the magic quadrant before pressing "submit?"

Next we come to IBM. They're certainly "mature," and I love that "you booted your old CEO out into the snow" is a strength.

It's pretty clear that @gartner_inc needs three Caution points for every entrant, so they had to expand this from its previous single-word entry: "Inertia."

Next up is @azure. I've seen board presentations that were less enterprisey than these three strengths.

The last two weaknesses speak directly to @ajassy's comments in @SiliconAngle today about SQL Server being obnoxiously priced everywhere but Azure. He said it way more eloquently, though.

No snark here. @OracleCloud has the best hybrid story by a lot: what you get in a public region, you can get on-prem.

I maintain @oraclecloud needs a story for folks with no Oracle databases at present.

This is very true. Oracle's reputation is that they'd steal a bone from a starving dog.

And then sue the dog.

"Their market share isn't great" seems like something that could apply to basically all of the players I didn't name and a few of the ones I did. Hmmmm...

So that's the high level overview. Now, my commentary:

Just who the hell evaluates a cloud provider's database offerings as the determining factor as to which cloud provider they're going to choose?!

I'm used to Route 53 being snubbed, but I bet there are a few other @awscloud service teams salty about being overlooked here.

I adore that despite nobody coming out SUPER well in this analysis, that doesn't stop Google, Amazon, Redis, and others from posting about this on their corporate blogs.

I will now take questions.

Everyone's database is just bits on a disk somewhere.

https://twitter.com/Nican/status/1333573448971345920?s=20

Hell yes it does! I'd want to have an exodus strategy from DynamoDB before building it as the core of my app's architecture.

@ScyllaDB is doing interesting stuff here, but I'd think long and hard before using DynamoDB for something I may need to move.

https://twitter.com/nicolaus/status/1333571741302595585?s=20

It's a good question. It's not immediately obvious to me what @MongoDB failed on from the list of inclusion criteria (global market presence, 3 DBMS use cases, reference-able clients in at least 3 industries from a list).

https://twitter.com/DorLaor/status/1333574363996590080?s=20

Nah. A routine capacity increase hit a hidden landmine limit. These things happen with computers.

https://twitter.com/tehsuck/status/1333570199946194945?s=20

If you've enjoyed this thread, look for more #reQuinnvent shenanigans this week, my own version of AWS's own version of Cloud Next. requinnvent.com

"Hope" is the most commonly used but least commonly cited strategy in all of cloud. @CockroachDB isn't a terrible option, but as with anything you've gotta know how it breaks.

https://twitter.com/StepanDavidovic/status/1333582237053366275?s=20

"Getting the work done" vs. "dicking around" is the perennial struggle. Different providers thus try different approaches to make getting work done sound like dicking around.

https://twitter.com/GBrayUT/status/1333578750013841408?s=20

I'm pretty sure you're me. Phrasing is super important apparently, as is adult supervision.

https://twitter.com/SecShea/status/1333577195223089152?s=20