Simonas Kazlauskas

I [wrote down my ideas](https://github.com/rust-lang/rust/issues/44489#issuecomment-367301641) on how we should fix the underlying issue. I still am going to write a separate issue for the underlying issue, of course.

FWIW it is probably a good idea to set up such services so that they attempt restarting themselves anyway. Especially things like sshd, a crash of which may lead to...

Given that most of these cases would be caught by crater, maybe its worthwhile to just make the change and make sure crates are fixed up? Alternatively: maybe only apply...

Note: this doesn’t actually add the API to measureme to output such data. Ideas on what API would make most sense here would be appreciated.

I’m probably not going to pursue this further, but if anybody wants feel free to pick this up.

We do have a `near-crypto` crate now, but is it being compiled to wasm32 and are people using it? EDIT: Ah, never mind, I didn't read the original issue statement...

https://buildkite.com/nearprotocol/nearcore/builds/23318#018512c9-e764-40f8-b2ba-bccf5ecb67eb is another flaky test (`tests::network::ban_peers::dont_connect_to_banned_peer`) specimen. [The logs from this test run (in case the link expires)](https://github.com/near/nearcore/files/10236699/nearcore_build_23318_cargo-test-integration-tests-star.log.gz)

Alternative is for homu to edit its own message and separate edits by horizontal rule. It would work because edits won’t send more mail and AFAIK pinging in edits will....

FWIW #26 was so long ago, I pretty much forgotten all about what it was about. That said, it isn't clear to me some of the premises here are accurate:...

So, I kept an eye on this and have found that the memory use grows slowly, over time. I have also found that the leak(?) occurs when the portal implementation...