JLockerman

@jgallagher We really appreciate any comments! Especially for tricky code like this 😊 I had (incorrectly) thought that the `nounwind` marker on the allocator code had been removed in anticipation...

Thank you for pointing this out!

The `oom=panic` PR was [recently merged](https://github.com/rust-lang/rust/pull/88098).

The unstable flag was merged as mentioned above, it's released unstably as of `1.6.1`, but is not yet stable ([tracking issue](https://github.com/rust-lang/rust/issues/43596)) we can either wait until it hits stable, or...

@mirandaauhl any thoughts about including this in `stats_agg()` versus creating a new aggregate for it? On the one hand, adding it to stats agg would keep it as a good...

> Now I guess I have a question haha. I'm curious what audience of engineers do we expect to use stats_agg()? Right now we're thinking of `stats_agg()` as a catchall:...

This needs communication with the core DB team; if we ever want to do the timebucket optimizations for this function as well, we'll really want it to be there. @davidkohn88...

Definitely looks like linker errors. I wonder if the environment variable `RUSTFLAGS` gets overridden by rustflags in the [`.cargo/config`](https://github.com/timescale/timescaledb-toolkit/blob/d173aa722408fa45240adc374267a43fa755699f/.cargo/config) during some passes, so we don't pass the ASAN arguments for...

(my gut wants to say that the second one is caused by postgres not being compiled with santizers, but I have no reason to actually think that...)

Partially done, on clients this works for TCP and needs a slight modification to work for UDP.