I am going to start posting more here. What kind of database content do people want?

- @planetscale.com by @samlambert.com: A MySQL-compatible serverless database with built-in scaling, branching, and non-blocking schema changes.

🔹 planetscale.com — Zero-to-Sharded in Minutes @samlambert.com at #sfruby live-sharded a MySQL cluster with zero application changes and it worked instantly. - Vitess handles ID generation and routing Watching “unsharded -> sharded -> serving traffic” in minutes felt unreal.

PlanetScale now supports PgBouncers (connection poolers) for your replicas. Connection pooling is broadly important for databases, but especially for Postgres because of its process-per-connection architecture. Don't know what that means? I have the perfect article!

Schema recommendations are available for Postgres databases. We analyze your database daily and provide recommendations for optimization: - Remove redundant indexes - Prevent primary key ID exhaustion - Drop unused tables and indexes - Rebuild bloated tables and indexes

This week on Mostly Technical, @aaronfrancis.com talks to @ianlandsman.com about how he reunited with @samlambert.com to launch his new YouTube channel and the incredible news he learned in the middle of emceeing Laracon.

Issue #337 of Off-by-none is out! In this issue, Amazon Nova Act brings AI agents to your IDE, Cloudflare backs Ladybird to push the open web forward, and PlanetScale for Postgres goes GA. #offbynone offbynone.io/issues/337/

PlanetScale for Postgres is now GA. planetscale.com/blog/planets...

13 Aurora clusters to a sleek move onto PlanetScale. Intercom’s database journey wasn’t just a migration, it was a rethink. On the latest Screaming in the Cloud, @bscanlan.bsky.social (Senior Principal Engineer @ Intercom) joins Corey Quinn to talk about..

- Why Aurora became operational pain - How PlanetScale’s partnership-driven model beat AWS’s building-block approach - Intercom’s pivot to AI agents post-ChatGPT - And why the shrinking pipeline of systems engineers is a growing concern. Listen 🎧: www.lastweekinaws.com/podcast/scre...

Listen to @bscanlan.bsky.social talking to @quinnypig.com about how Intercom moved from AWS Aurora to @planetscale.com t.co/ZlRtVVr7Rs

Database summary card shown on @planetscale.com dashboard. Metrics link to corresponding pages. This example has two keyspaces, the sharded one running on NVMe Metal storage.

Anyone want to jump on a zoom to import their postgres database to @planetscale.com? We've got a coupon $100 for you

Will it recommend moving to @planetscale.com instead? Because based upon most of the analyses I've run, it probably should.

we have a postgres -> PlanetScale import tools that are ready for release soon. if you want some while glove help using them while we test, email me: [email protected]

$20K/month is where your AWS bill changes the most. •$6K on RDS IOPS. •$4K on transit charges between AZs you forgot •$3K on NAT gateways passing idle traffic •$2K on S3 PUTs from the microservice the principal engineer wrote •$5K on CloudWatch logs

Vectors + Metal is an insane combo. We made vector storage and search into MySQL meaning PlanetScale is the first RDBMS to allow vector search as part of a full SQL query. You can pre and post filter while doing vector lookups reducing 3-4 queries into 1

Intercom saw a significant improvement in both tail latency and hardware cost reduction after switching to Metal. “PlanetScale Metal has been a game changer for Intercom. All of our biggest, most critical PlanetScale databases now run on Metal.” Read their story here: pscale.link/int

Based on price, performance, and (ahem) reliability, I’m really puzzled as to who’d pick Aurora over PlanetScale. Am I missing something obvious?

Seven days after I joined @planetscale.com, I pitched what would become PlanetScale Metal to @samlambert.com. Now all of PlanetScale's customers can run Vitess the way Slack runs Vitess, on the fastest NVMe drives you can get in the cloud. planetscale.com/blog/planets...

I've been helping the open source Vitess team with a rebrand, starting with an updated logo. The mark represents horizontal database sharding. Hopefully a new site design will follow soon.

I love monorepo build tools because I want to constantly be on the verge of walking into the sea.

how do people not understand that wire protocol compatibility is a different thing from syntax, behavior, and data format compatibility?

it’s incredibly difficult for a male over the age of 35 to wear a baseball cap without looking like a toddler

whenever you log onto a new server you have to run ‘w’ a few times to let it know you are there. just like patting a horse on its head.

"now we are at scale i am really glad we rely on stored procedures and triggers" - nobody ever

its insane that there isn't a new apple mouse with magsafe

i’m not actually in it for the christmas dinner. for me it’s all about the leftovers sandwiches.

make your wife breakfast

for the first time in my life i am considering buying an android phone

in the next 30 years there will be a mental health crisis because of the realization we are a few generations too early for immortality and gigantic technological change

the academic/theoretical side of the database community is obviously valuable, but most of the debate is irrelevant to how databases run in production

meetings

i love my job and the people i work with

do you understand why shared nothing is the ideal architecture? if not, would you like to?

a key part of onboarding customers is the joint sharing of slack emojis

99% of benchmarking is a complete waste of time

i can save you some time and tell you that nobody wants to read your 2024 year in review blog post

Cash App's move to PlanetScale required migrating over 400 shards - all with no interruption to the millions of daily transactions and zero impact on customers.

every tech company has a totally cracked systems team fighting to claim back 100ms of page load time just so a product team can eat it back up with the worst feature you can possibly imagine

its wild that people believe architectural decisions that are good for OLAP are default good for OLTP

designers love it when you tell them they should "make it pop"

faster databases = less need for front end capacity

gmail is being used for spam at a scale ive not seen before

chatgpt real time video is fucking incredible

it feels nice to know the cloud database you built with your friends hosts products with over 200,000,000 combined MAU

broke: use devin woke: clone your best engineers

fast

the real monolith is the database

you can now configure how we throttle your schema changes planetscale.com/changelog/th...