John Howard
John Howard
this is done ``` $ cargo tree | rg hyper ├── hyper v1.4.1 ├── hyper-rustls v0.27.2 │ ├── hyper v1.4.1 (*) │ ├── hyper-util v0.1.7 │ │ ├── hyper v1.4.1...
I am seeing this even with a single connection that sends a large amount of data. so it may be some limit but I don't think it would be num...
fixed by https://github.com/istio/ztunnel/pull/1067
PR merged so hopefully they tag a release. unfortunately you need a `patch` in cargo if not, unlike go
https://github.com/polarsignals/rust-jemalloc-pprof or https://github.com/mkmik/heappy/ seem viable. Prototype with heappy: https://github.com/howardjohn/ztunnel/pull/new/exp/heappy
I agree with Nate On Thu, Nov 9, 2023 at 10:14 AM Nathan Mittler ***@***.***> wrote: > @zhlsunshine just to be clear, you > haven't actually seen performance problems in...
One thing to note is the behavior currently is equal to without DNS proxy. So it's not like this is fixing a regression, purely a performance benefit , and the...
https://github.com/istio/ztunnel/pull/1098
Deduping with https://github.com/istio/ztunnel/issues/214 (which has some more comments)
Tokio has some 'budget' concept but it is per-task. We would want some concept of task-groups. But we also probably want higher level fairness like bytes/s https://docs.rs/tokio/latest/src/tokio/runtime/coop.rs.html