hyper 0.12 benchmark results place it at 99%: https://www.techempower.com/benchmarks/#section=test&runid=a7f29eb8-862b-49ce-9789-ff72a6410800&hw=ph&test=plaintext … The majority that performed better are less complete or unsafe, so I think that's pretty good!
-
-
Replying to @seanmonstar @steveklabnik
I just glanced at the plaintext benchmark, and the latency appears to be 1.4ms versus Actix(not Actix-raw) hitting 0.3ms. Do you know why such a big difference?
1 reply 0 retweets 0 likes -
Replying to @reacharavindh @steveklabnik
I haven't investigated. I know one other difference in that benchmark that can fluctuate hyper's results is that it doesn't balance connections equally on all cores, relying instead on the kernel. That means some runs can have a core with more connections than others.
1 reply 0 retweets 1 like
Though to note, this is behavior of that specific benchmark, not a requirement of hyper. Perfect round robin is best for that benchmark, since the connections persist and have the same load, but it's not realistic of the real world.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.