More than a refresh: Conversations with the most interesting people you've never met
Oct 15, 2026 · 20 min · 7 segments
In Part 2 of "When Postgres Is Enough", Michael Cooper and JD examine what it took for AgLedger to trust PostgreSQL with work previously split between PostgreSQL and Valkey. Supporting two databases…
You've actually reached or kind of accidentally fell into Postgres' sweet spot.
So one of the things I talk to people about, you always hear, well, I can't do this.
It's like, yeah, but literally 1% of the planet, if that, needs that, right? Postgres has this very broad sweet spot.
And I'm not saying it's faster than everything, but almost always it can either be tuned or engineered to get enough so you don't need something else.
Right? And that's something a lot of people
And we can talk about that in Series 3 because, again, I refreshed a lot of my testing last night against the latest Postgres.
And so... all of the tuning I was doing was just, you know, how can I get 20% more off of a box that again, it's costing 60 bucks an hour.
And you know, my ICP is never going to be running my instances on, on the size of boxes I'm testing on.
And so to find, I, I totally agree there are contested points and I do, I do believe there's, you know, some bloat and issues along the way.
And we can, and in, in, in series three, we can also talk about PG boss and some of the things that we were doing there.
Well, and even with that, there's the other side of it is that I have found that what a lot of people will do is say Postgres is slow at a microservice or at this other managed service because they're not database people.
Right? A database person would say, whoa, no, hold on.
Let's figure out what's going on underneath.
Before we add anything, because I can bet you dollars to donuts, either if I add some RAM or I clean up some bloat or I change this parameter, it would be better parallelized.
You've actually reached or kind of accidentally fell into Postgres' sweet spot.
So one of the things I talk to people about, you always hear, well, I can't do this.
It's like, yeah, but literally 1% of the planet, if that, needs that, right? Postgres has this very broad sweet spot.
And I'm not saying it's faster than everything, but almost always it can either be tuned or engineered to get enough so you don't need something else.
Right? And that's something a lot of people
And we can talk about that in Series 3 because, again, I refreshed a lot of my testing last night against the latest Postgres.
And so... all of the tuning I was doing was just, you know, how can I get 20% more off of a box that again, it's costing 60 bucks an hour.
And you know, my ICP is never going to be running my instances on, on the size of boxes I'm testing on.
And so to find, I, I totally agree there are contested points and I do, I do believe there's, you know, some bloat and issues along the way.
And we can, and in, in, in series three, we can also talk about PG boss and some of the things that we were doing there.
Well, and even with that, there's the other side of it is that I have found that what a lot of people will do is say Postgres is slow at a microservice or at this other managed service because they're not database people.
Right? A database person would say, whoa, no, hold on.
Let's figure out what's going on underneath.
Before we add anything, because I can bet you dollars to donuts, either if I add some RAM or I clean up some bloat or I change this parameter, it would be better parallelized.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.