This blew me away. I have been using AWS in production for ten years. Until today, I made the assumption that their Relational Database Service (RDS) carried around a 33% premium over EC2. That’s the way it started out. AWS usually drops prices over time, right?
RDS uses the same MySQL and PostgreSQL software available for download to the public. Other than that, it’s basically a bunch of well-oiled provisioning, failover, and backup scripts¹. The tacked on support isn’t life-changing, to say the least. But a 33% premium seems pretty reasonable to avoid some drudgery. After all, what is my time worth?
But a 78% premium? Now that gives me pause. Particularly because it is so much more and with no added value². This violates a core expectation of technology: it gets better and cheaper over time even if it was already worth it.
Admittedly I am quite late to the party on this one. I couldn’t find the M4 announcement for RDS, but generally it takes about a year for new instance types to percolate from EC2 to RDS. That puts it at roughly 2016. I’m three years late.
I know, I know. AWS is not actually for the price sensitive, even though every one of their product pages talks about how “inexpensive” it is. It never has been. Their target market is those who prioritize time-to-market above all other concerns³. I get it.
But I don’t want to feel stuck⁴. When a price suddenly doubles, I immediately think to myself “is this still the right choice?” I usually come to the conclusion that I don’t actually have a choice and I get out the wallet. Then I begin to feel like I’m under the vendor’s thumb. I feel betrayed.
That’s the way I feel about comic-book-villain-esque companies of yore like Oracle and Microsoft. I start to wonder if AWS is still hungry. I start to wonder if I should continue to deepen my personal investment in their platform. I start to wonder if I should still recommend them as the obvious choice.
Technology should get better and cheaper over time even if it was already worth it. Companies that operate by this principle have my allegiance. Companies that do otherwise don’t deserve my business.
 Aurora is a very different beast, but that’s for another time.
 Upgrading from M3 to M4 is a paltry increase in compute power. I know there are a few other small performance-oriented improvements like increased EBS bandwidth. Let’s be real though, this is not commonly a key performance driver for a relational database. CPU, Memory, and IOPS are what really make the difference.
 You may ask: “what business doesn’t put time-to-market first???” To that I say: in my experience, most of them!
 Even if it might already be the truth! Human decision-making is largely emotional, not logical. I have no evidence that I am an exception.