Aaron

@yliu-intuit: I just saw your video and am kind of confused at what might be wrong; it doesn't look like the config (from the logs, I think it's being taken...

@yliu-intuit: ahhh! yeah, that'd do it. I forgot that I missed the cut date for 2.9.0 with that pr; it should make it into 2.10.0, which should be sometime this...

@montykamath: thanks for the repro! this is super interesting and we're looking at it (alongside other investigations into v2.4.0 that line up with this)

@montykamath: no significant progress yet! I'm going to spend some time today with your draft pr to see what we can learn from it; it'd be a much longer, broader...

hey @montykamath: we've identified the core problem and have a two-pronged approach we're experimenting with (one for router-side mem safety and one for federation-side mem safety; the latter is important...

@montykamath: we're still making progress! it's slow because we have other customer interrupts; so, it'd actually be super helpful if you'd reach out to either your account rep or open...

hi @aditi-rajawat! the 16kb is a [default set by hyper](https://github.com/hyperium/hyper/blob/master/src/proto/h2/server.rs#L40) and we don't yet expose a way to configure that, but we're actively thinking about how we might expose it....

hey @aditi-rajawat! it's in the right direction; we're working on something similar, and will push forward with what we have to get it out a little faster (ie, reduce the...

@aditi-rajawat: I was OOO at the end of last week, so replying here (instead of the closed PRs) about the tls versus tcp connections: I haven't looked at your new...

@aditi-rajawat: a fix for tcp/uds should be merged today or early next week and make it into the next release (2.10, I believe)