Craig
@craigkerstiens.com
Product @crunchydata previously MSFT/Citus/Heroku. Talk a lot about Postgres and startups. Why Postgres? https://www.crunchydata.com/why-postgres
The details matter. In this case, completely revamping our newsletter signup screen ahead of an upcoming conference.
I mean why wouldn't you execute SQL to sign up for a database newsletter?
I mean why wouldn't you execute SQL to sign up for a database newsletter?
May 7, 2025 at 5:52 PM
The details matter. In this case, completely revamping our newsletter signup screen ahead of an upcoming conference.
I mean why wouldn't you execute SQL to sign up for a database newsletter?
I mean why wouldn't you execute SQL to sign up for a database newsletter?
@andypavlo.bsky.social on HN sums it up well
OLAP vs. OLTP isn't right vs. wrong, they're designed for different things.
Have a version of this on a slide when explaining Crunchy Data Warehouse and how it's for very different purposes than stock Postgres.
OLAP vs. OLTP isn't right vs. wrong, they're designed for different things.
Have a version of this on a slide when explaining Crunchy Data Warehouse and how it's for very different purposes than stock Postgres.
April 22, 2025 at 10:40 PM
@andypavlo.bsky.social on HN sums it up well
OLAP vs. OLTP isn't right vs. wrong, they're designed for different things.
Have a version of this on a slide when explaining Crunchy Data Warehouse and how it's for very different purposes than stock Postgres.
OLAP vs. OLTP isn't right vs. wrong, they're designed for different things.
Have a version of this on a slide when explaining Crunchy Data Warehouse and how it's for very different purposes than stock Postgres.
🐘 meets 🧊
End to end, under 2 minutes.
- Two commands to replicate data from Postgres -> Iceberg
- Synced over 10m rows under a minute
- Data is continually processed and updated in Iceberg
- count(*) in Postgres over 300ms down to under 20ms
End to end, under 2 minutes.
- Two commands to replicate data from Postgres -> Iceberg
- Synced over 10m rows under a minute
- Data is continually processed and updated in Iceberg
- count(*) in Postgres over 300ms down to under 20ms
April 22, 2025 at 2:38 PM
🐘 meets 🧊
End to end, under 2 minutes.
- Two commands to replicate data from Postgres -> Iceberg
- Synced over 10m rows under a minute
- Data is continually processed and updated in Iceberg
- count(*) in Postgres over 300ms down to under 20ms
End to end, under 2 minutes.
- Two commands to replicate data from Postgres -> Iceberg
- Synced over 10m rows under a minute
- Data is continually processed and updated in Iceberg
- count(*) in Postgres over 300ms down to under 20ms
What's this? Flexible/variable long term backup retention 👀
April 8, 2025 at 9:18 PM
What's this? Flexible/variable long term backup retention 👀
Once you get the basic concepts a lot of power is suddenly unlocked.
April 4, 2025 at 4:38 PM
Once you get the basic concepts a lot of power is suddenly unlocked.
I've encountered a database without constraints that "enforces them within the app" I've encountered data quality issues. Postgres has a variety of constraints to help, not just primary/unique and foreign keys, but also not null and check constraints.
March 27, 2025 at 4:08 PM
I've encountered a database without constraints that "enforces them within the app" I've encountered data quality issues. Postgres has a variety of constraints to help, not just primary/unique and foreign keys, but also not null and check constraints.
Something I've always found pride in is making developer tooling more accessible. You won't find many that think Postgres isn't powerful, but navigating it can be overwhelming. Scheduled jobs in a database aren't a "new thing", pg_cron has long been around but it wasn't always "friendly".
March 26, 2025 at 6:31 PM
Something I've always found pride in is making developer tooling more accessible. You won't find many that think Postgres isn't powerful, but navigating it can be overwhelming. Scheduled jobs in a database aren't a "new thing", pg_cron has long been around but it wasn't always "friendly".
Partitioning in Postgres is great if you've got any form of time series data. It can allow higher ingest, faster queries, and easily archiving old data. Common examples could be:
- logs
- events
- metrics
- logs
- events
- metrics
March 25, 2025 at 6:31 PM
Partitioning in Postgres is great if you've got any form of time series data. It can allow higher ingest, faster queries, and easily archiving old data. Common examples could be:
- logs
- events
- metrics
- logs
- events
- metrics
Can you guess when we moved internal logging from cloudwatch over to S3/Iceberg via Crunchy Data Warehouse?
March 21, 2025 at 4:55 PM
Can you guess when we moved internal logging from cloudwatch over to S3/Iceberg via Crunchy Data Warehouse?
Spent a little time this weekend working on some temporary tattoos for the kids. Very curious to see what they end up thinking of them.
March 17, 2025 at 4:03 PM
Spent a little time this weekend working on some temporary tattoos for the kids. Very curious to see what they end up thinking of them.
What you're vendor tells you doesn't *always* align with reality, it aligns with their incentives.
This aligns with what I've seen that typically "vanilla" RDS is often an improvement over Aurora.
This aligns with what I've seen that typically "vanilla" RDS is often an improvement over Aurora.
March 14, 2025 at 8:24 PM
What you're vendor tells you doesn't *always* align with reality, it aligns with their incentives.
This aligns with what I've seen that typically "vanilla" RDS is often an improvement over Aurora.
This aligns with what I've seen that typically "vanilla" RDS is often an improvement over Aurora.
A big part of building Crunchy Data Warehouse was ease of use. How easy is it to load data from existing public datasets?
Step 1: Point at your dataset and we'll load it for you
Step 2: Query it
Step 3: Profit
Step 1: Point at your dataset and we'll load it for you
Step 2: Query it
Step 3: Profit
February 27, 2025 at 6:39 PM
A big part of building Crunchy Data Warehouse was ease of use. How easy is it to load data from existing public datasets?
Step 1: Point at your dataset and we'll load it for you
Step 2: Query it
Step 3: Profit
Step 1: Point at your dataset and we'll load it for you
Step 2: Query it
Step 3: Profit
Are PMs folks that can actually get things done?
January 29, 2025 at 8:49 PM
Are PMs folks that can actually get things done?
Picked up a few solid bottles today
January 11, 2025 at 1:51 AM
Picked up a few solid bottles today
In addition to usual butcher box picked up some bones as extra. Beef stock for (maybe) the month
January 8, 2025 at 7:07 PM
In addition to usual butcher box picked up some bones as extra. Beef stock for (maybe) the month
Feedback from users can be rare, especially when you lose a customer. Take pride in your work and run a good service, and hope one day for users like this.
(Context: Their side project shut down, and they sent a personal note saying sorry they're no longer a customer).
(Context: Their side project shut down, and they sent a personal note saying sorry they're no longer a customer).
December 30, 2024 at 4:49 PM
Feedback from users can be rare, especially when you lose a customer. Take pride in your work and run a good service, and hope one day for users like this.
(Context: Their side project shut down, and they sent a personal note saying sorry they're no longer a customer).
(Context: Their side project shut down, and they sent a personal note saying sorry they're no longer a customer).
If you're brand is be informative and helpful. You literally print stickers that contain Postgres tips on them.
December 27, 2024 at 4:58 PM
If you're brand is be informative and helpful. You literally print stickers that contain Postgres tips on them.
A while back we added a command palette to Crunchy Bridge (⌘ + K and get a set of shortcuts that you could easily navigate by typing). I personally love and use it all the time!
BUT, most of our users don't log in every day. So we exposed it directly in the nav to make it easy to find.
BUT, most of our users don't log in every day. So we exposed it directly in the nav to make it easy to find.
December 16, 2024 at 7:20 PM
A while back we added a command palette to Crunchy Bridge (⌘ + K and get a set of shortcuts that you could easily navigate by typing). I personally love and use it all the time!
BUT, most of our users don't log in every day. So we exposed it directly in the nav to make it easy to find.
BUT, most of our users don't log in every day. So we exposed it directly in the nav to make it easy to find.
At movie theatre power went out, unsure if they’ll be able to play movie they make announcement about it.
Kid says can we start the movie back.
I ask if he was listening.
Apparently he meant this movie
Kid says can we start the movie back.
I ask if he was listening.
Apparently he meant this movie
December 14, 2024 at 6:26 PM
At movie theatre power went out, unsure if they’ll be able to play movie they make announcement about it.
Kid says can we start the movie back.
I ask if he was listening.
Apparently he meant this movie
Kid says can we start the movie back.
I ask if he was listening.
Apparently he meant this movie
Not a fruit, but availability of fun produce in this area is so great
December 12, 2024 at 2:57 AM
Not a fruit, but availability of fun produce in this area is so great
Weekends are for broths and stock
December 9, 2024 at 2:19 AM
Weekends are for broths and stock
CREATE TABLE measurements_staging (like measurements);
INSERT INTO measurements_staging VALUES ('Charleston', 21.3);
WITH new_rows as (
DELETE FROM measurements_staging returning *
) INSERT INTO measurements SELECT * FROM new_rows;
INSERT INTO measurements_staging VALUES ('Charleston', 21.3);
WITH new_rows as (
DELETE FROM measurements_staging returning *
) INSERT INTO measurements SELECT * FROM new_rows;
December 5, 2024 at 5:06 PM
CREATE TABLE measurements_staging (like measurements);
INSERT INTO measurements_staging VALUES ('Charleston', 21.3);
WITH new_rows as (
DELETE FROM measurements_staging returning *
) INSERT INTO measurements SELECT * FROM new_rows;
INSERT INTO measurements_staging VALUES ('Charleston', 21.3);
WITH new_rows as (
DELETE FROM measurements_staging returning *
) INSERT INTO measurements SELECT * FROM new_rows;
It's @winsletts.bsky.social who is on the front of the bow isn't it
December 4, 2024 at 8:13 PM
It's @winsletts.bsky.social who is on the front of the bow isn't it
"Postgres compatible"
- No views/triggers/sequences
- No foreign key constraints
- No extensions
- No NOTIFY ("ERROR: Function pg_notify not supported")
- No nested transactions
- No JSONB
What, what IS it compatible with?
- No views/triggers/sequences
- No foreign key constraints
- No extensions
- No NOTIFY ("ERROR: Function pg_notify not supported")
- No nested transactions
- No JSONB
What, what IS it compatible with?
December 3, 2024 at 7:31 PM
"Postgres compatible"
- No views/triggers/sequences
- No foreign key constraints
- No extensions
- No NOTIFY ("ERROR: Function pg_notify not supported")
- No nested transactions
- No JSONB
What, what IS it compatible with?
- No views/triggers/sequences
- No foreign key constraints
- No extensions
- No NOTIFY ("ERROR: Function pg_notify not supported")
- No nested transactions
- No JSONB
What, what IS it compatible with?
Managed to dig up a few of the pages as well
December 3, 2024 at 4:27 PM
Managed to dig up a few of the pages as well