RDS vs. PlanetScale: Pricing Considerations | Hacker News Reader

RDS vs. PlanetScale: Pricing Considerations

vantage.sh

78 pointssamlambert17 comments

Cool, one poorly optimised query that scans a 500 million row table will cost you 75$ for the scan alone. Assuming a 100gb table that’s an additional $125.

Per. Query.

We have features coming that will make this not a problem for people.

is that feature called - automatic query optimization?

Awesome! I like the idea of the product, but the pricing structure is really really risky and makes it a hard sell. At least with AWS the price is fixed instead of “possibly cheap or possibly the most expensive thing you can ever imagine”.

It’s hard to know how you’d get rid of this problem: either you block terrible queries (bad for services) or have some price caps/forgiveness (bad for you).

Maybe I am being cynical here, but one advantage of AWS is that when it goes down, the whole online world looks like it is on fire.

So if you have a database on AWS, and AWS goes down, you likely don't even need to tell your CEO that AWS is down, they already probably know that from the news or because their favorite website/app is also down.

However, if PlanetScale goes down, likely everything else appears to be working except yours.

I don't think this is cynical, it's a restatement of the "nobody gets fired for buying IBM" adage. Choosing PlanetScale at this stage in its market penetration is almost certainly going to be seen as the riskier business choice over RDS Aurora (even though it's not quite apples:apples) or GCP Spanner (which is). You can choose to spend your innovation token there, but you better be prepared for the downside when PlanetScale goes down and everything else is up.

I love the Innovation Token blog post, but. it's been a while since I've seen the OG post. Do you have a link?

http://boringtechnology.club/

I have my problems with this statement mostly because IBM actually does suck. The amount of products its bought and promptly did a full thrust crash into the ground are many. AWS has its problems but 99.99999% of the time the "core" products perfectly fine, I dont know about its multitude of side-hustles though.

I think you're missing the context that the adage "nobody gets fired for buying IBM" dates back to the 1970s at least. It's not meant to refer to IBM as it is now, but rather to IBM as it was then, an industrial behemoth with a near monopoly on business machines, some of which happened to be general purpose computers.

IBM used to be good

PlanetScale probably runs on AWS or GCP, so that’s a solved problem (:

I've been eyeing PlantScale since they went GA. I pay $45.86/mo (or at least that's what I paid last month, thanks Vantage!) for Aurora Serverless scaled down to 1 instance (this is my prod server, my dev/qa will shutdown after 5min of inactivity so they cost near $0). My traffic is very bursty (as in no traffic at all then for 2 days it ramps up then back down to nothing, it's software for festivals) so "Serverless" was very attractive to me as I don't want to deal with the bursts even if it costs me more on the off-months/days. PlanetScale looks very cool and I'm in the process of moving off the data-api for Aurora to an ORM in part so that I could swap the underlying DB easier. I also like the concepts of branches that PlanetScale offers.

My biggest hangup with PlanetScale is the unknown cost. Maybe I'm a bad dev or I should have better stats but I have no idea how many rows are read/written (I can get the second easier of course, or at least the initial write and I don't update much beyond the profile table). Runaway bills are terrifying for a solo dev and at least with Aurora I was able to calculate what it would cost to run my prod server at full (256 instances I think? or is 128?) for the duration of the festival and I was ok with that cost if it came to that. PlanetScale is somewhat limitless and I can't have the DB go down in the middle of a festival if I hit a spending limit.

For all I know my usage will fall 100% within the "base tier" ($30/mo) but it's really hard to know (or at least it is for me). I might just need to bite the bullet and try them but it's not a zero-cost to switch DBs (I'm using MySQL at least on Aurora as well).

Interesting, a more direct comparison is probably Aurora and they seem to have similar pricing.

Hey there - I'm the author of the post. It's a really good call out and something we're thinking about as a future post in the series.

When we took a look at RDS costs, the vast majority of customers are using "raw" RDS. So the hope with doing this analysis (which was definitely a bit trickier to do) was to try and purposefully compare that with PlanetScale as that seems to be the "true" customer anchor.

If you are measuring a monthly price in 10ths of a penny - the cost can't be that high. Does AWS even bill monthly in tenths of a penny? I normally calculate AWS projections in dollars.

I think folks are used to RDS type pricing models, so would expect that to remain popular.

Some notes:

Reserved instance pricing easily saves 40%+ on a one year term with AWS. For folks where RDS is really at the top end of their budget, I think its a bit weird not to bring costs down with reserved instances - most large production workloads don't handle being spun up and down every hour etc. Not sure what planetscale offers there.

For dynamic pricing AWS offers other datastores. Aurora Serverless 2 is I think

$0.10 per GB/month of storage. So $100K per PB? $0.30 per million transactions

They say

"Whatever that number produces you can line up with the Amazon RDS multi-AZ price you’re currently paying Amazon."

Do we have confirmation that plantscale has a proper multi-AZ architecture?

But...MySQL? Sigh...

I migrated to PostgreSQL later than most, and I've still be using Postgres for six years...

Reply on news.ycombinator.com