Joonas Koivunen

thought about this a bit more, updated the description, marking this draft.

> > overall this is insufficient because we have no signal when the basebackup requester dies even a single tenant with longer apply time than retry basebackup time could oom...

Noting on #6628 why I'll most likely close this.

I haven't noticed this, most of the time recompilations I assume happen because of postgres v14 is using blindly overwriting install... Let's keep this open, I'll try to notice this...

I don't understand how this is a problem, does it happen because we are crossing the crate boundaries? https://play.rust-lang.org/?version=stable&mode=debug&edition=2021&gist=4809433abb7443bc9c7d7f5bde8b4756

Discussed about this, the issue is about that the fact that `(B, Result)` return type allows *getting the buffer* while not getting the warning on the result, like: 1. `let...

Yes, forgot to link it with the PR... Well, I'll check in tomorrow actually.

The condition after #6661 are: 1. distance from last is enough 2. size would be greater than limit 3. non-zero distance AND enough time As such I don't see how...

I should had edited that :) > Yeah, I agree. The issue still stands. > > > How to fix it in my mind, compare first write to an open...

> If we actually wanted to respect the checkpoint timeout, then we'd have a background task that wakes periodically and requests a flush if the last freeze was too long...