Postgres FM · Hands-On Engineering Podcasts · November 2026
The hosts were discussing Postgres 19's new plan advice module, which stops short of traditional query hints. Kirk Wolak explains why hints baked into application code age badly, and why he'd rather pin a good plan at the database level.
We
…methodology, but like without any gaps, but it's quite like expensive to use. You need to actually allocate resources to study all the plan flips and then what to do with flips. Now we have a solution what to do with flips, but still plan flip analysis remains another challenge, right?
…literally a slash asteris hint after the word select on my thing. And I'm going to put some things in there and tell you what tables and indexes you should really be doing as you're either your driver table or primary, whatever. I'm going to focus you because you're going to make mistakes. And the problem is, I want you to hold it right there. It was a great idea in 1975. Okay. The problem is. We're 2026, baby. And the reality is you're hard coding the hint usually into the application. The last thing you want to do is be tweaking your application because of a subtle change in the database, because now your hint is actually a bad idea. And that happens. And this plan with stashing works differently. And what the high-level version is this: you can actually calculate a good plan, and then you can associate it with the SQL ID. Okay. And then you can give it priority. And this gives you a chance as the DBA to change the behavior of the application without getting the application developer involved. He can ring you up and say…
As I see, our testing found some overhead from using advice. Yeah, like around 10%, right? So it's not for free when we talk about very fast queries. So, anyway, there are some long list of limitations, but it's a great step in an interesting direction, which I think bigger databases actually need.…
I have one that's. You got one more?
Go ahead.